← All docs

Legal analysis — State ex rel. Parent v. Allen County Board of Commissioners

Document: allen-co-bosc-mandamus-draft-2026-05.pdf (draft Complaint for Writ of Mandamus, Court of Appeals of Ohio, Third Appellate District, Allen County; Relator Cory Parent, pro se; Certificate of Service dated May 31, 2026). Drafted but not filed — Relator received the Clerk’s acknowledgment (Ex. E, Apr 29) and elected to hold the complaint.

Analysis, not legal advice. This is a research memo over a primary source in the BOSC corpus; it characterizes the law as of June 2026 from public sources and does not verify full case texts. Statutory scope below is contestable. Relator is a competent pro se litigant; nothing here is a substitute for his own judgment or retained counsel.

0. What the document is

A public-records mandamus under R.C. 149.43(C)(1)(b) — the Public Records Act’s “original action in mandamus” track — seeking to compel the Allen County Board of Commissioners to produce records responsive to three requests in the BOSC campaign:

The relief sought: a writ compelling production + a withholding log, court costs, attorney’s fees (conditioned on later retaining counsel), and statutory damages of $100/business day to a $1,000 cap under R.C. 149.43(C)(3).

This is a well-constructed pro se filing. It does the three things that most pro se public-records complaints get wrong, and it anticipates the county’s single strongest defense before the county has raised it.

I. The case on its own terms

1. Route choice — (C)(1)(b) mandamus, not the Court of Claims

R.C. 149.43(C)(1) gives a requester two mutually exclusive paths, and the complaint (¶4) consciously elects (C)(1)(b) and disclaims (C)(1)(a):

• (C)(1)(a) — Court of Claims. $25 filing fee, mandatory mediation, a special master, statutory damages available — but no attorney’s fees, and it is slower. The cheap, low-friction path most pro se requesters take.
• (C)(1)(b) — mandamus in common pleas, the court of appeals (original jurisdiction, Ohio Const. art. IV §3), or the Supreme Court. Attorney’s fees + court costs + statutory damages are all available; no mediation gate.

Choosing the court of appeals is the aggressive, higher-leverage option. For a pro se relator with no counsel, the fees lever is dormant (you cannot recover fees you did not incur — hence ¶b’s “should he later retain counsel”), so the immediately recoverable relief is costs + the $1,000 statutory-damages cap. The strategic value of (C)(1)(b) is therefore less about money than about (a) speed and a real judicial forum rather than AG mediation, and (b) preserving the fees claim if the matter escalates and counsel appears. Sound choice given the subject matter.

2. The (C)(2) written affirmation — present and correct

Section V of the complaint is the R.C. 149.43(C)(2) affirmation: it states the request was served on the records custodian and that three business days elapsed before suit. This is not boilerplate — the statute mandates dismissal if it is omitted. Many pro se mandamus actions die on exactly this defect. Including it (with the May 26 cure notice as the predicate) is the filing’s most important piece of procedural hygiene.

3. Merits — clear right, clear duty, constructive denial

Public-records mandamus requires a clear legal right to the records, a clear legal duty to produce, and proof by clear and convincing evidence. The duty attaches once the office has fair notice of a request. On these facts:

• Fair notice is not seriously contestable — the county acknowledged receipt in writing (Ex. E) and described the requests with specificity (“four (4) records requests”).
• Unreasonable delay = constructive denial. R.C. 149.43(B)(1) requires records “within a reasonable period of time”; prolonged, unexplained silence is treated as a denial that mandamus will reach. The complaint’s theory (¶11) — that failure to produce, to provide a schedule, or to furnish a withholding log within a reasonable time is itself the §149.43(B) violation — is the correct framing.
• “Referral to counsel does not toll.” The cure notice’s central legal point is right and well-settled in spirit: an office cannot park a request with its lawyers indefinitely and call the clock stopped. Acknowledgment is not production.

The timeline is favorable. By the May 31 certificate date, ~6 weeks had passed since the April 14 request and ~4½ weeks since the April 29 acknowledgment, with zero records, zero schedule, and zero withholding log. Compare State ex rel. Henderson v. Washington Court House, 2026-Ohio-110, where four months of silence supported the action (the writ was ultimately mooted by production — see §I.5).

4. Damages math

R.C. 149.43(C)(2)/(3): statutory damages accrue only where the request was made by hand delivery, electronic submission, or certified mail — the requests here were email, which qualifies — at $100 per business day running from the date the office failed to comply, capped at $1,000 (i.e., ten business days). On this timeline the cap is already reached; the damages claim is essentially fixed at the $1,000 ceiling. Modest, but real, and it is a per-request entitlement in principle — worth noting the complaint pleads three requests.

5. The mootness problem — why holding the complaint is defensible, and its limit

The reason the complaint was not filed is the right instinct for a different reason than it might appear. Production moots the writ itself — once the county hands over the records, a court will not order it to do what it has already done (the exact result in Henderson, 2026-Ohio-110). But two things survive production:

1. Statutory damages and costs are not mooted by late production — they are the remedy for the period of non-compliance that already occurred. Henderson and the broader line (e.g., the pro se $1,000 award reported in DiFranco-type cases) keep the damages question live even after the records arrive.
2. A withholding log for anything actually withheld (esp. under §9.66(D), below) is independently compellable.

So the strategically precise posture is: holding the complaint while the county produces is reasonable; holding it while the county merely acknowledges and delays forfeits accruing leverage. The cure notice (Ex. F) correctly converts the acknowledgment into a dated trigger. The complaint was ripe on May 31. If the “confirmation of records requests” that prompted the hold was only a further acknowledgment (not production), the §149.43(B) violation persisted and the $1,000

• costs claim was already vested.

Confirmed outcome (per the 6/4 testimony). The leverage thesis is no longer hypothetical. Testifying to the Select Committee on June 4, the relator stated that the petition “is actually what finally got me those records on Friday, or at least a production schedule to produce them.” So the draft worked as leverage without being filed — the credible, court-ready threat broke a stonewall that mere cure-notice correspondence had not. That is the strongest possible vindication of §II below: the value of a complete, properly-pleaded mandamus is not only the writ it could win but the production it extracts before any judge sees it. (It also means a filed action would now likely face the Henderson mootness bar on the writ itself — while the statutory- damages/costs question for the pre-production delay remains live.)

6. Two cracks worth noting

• Date — reconciled to April 14. The operative date is April 14, 2026: the Ex. A Gmail transmission header reads Tue, Apr 14, 2026 15:31 -0400, and for public-records purposes the date of service (transmission) is what starts the clock, not a draft date. The “April 13” references in the first supplement (Ex. B) and the cure notice (Ex. F) are scrivener errors carried from the drafting day; they should be corrected to the 14th in the filing itself before it is used. (This is the date adopted in the committee testimony, and throughout this memo.)
• Scope mismatch. The county’s acknowledgment counts four requests; the complaint litigates three. The Downloads/Requests/ campaign in fact spans more (zoning amendment #6172-2026, township SIU communications, a moratorium request). Pin the universe of requests the writ covers, or the respondent will argue the pleading is under-inclusive/ambiguous.

II. The General Assembly dimension — the §9.66(D) headwind

The most consequential thing in the filing is buried in the cure notice (Ex. F): the R.C. 9.66(D) argument. This is not a public-records-mechanics issue; it is the substantive exemption that now shrouds exactly this category of records statewide, and it is the county’s strongest available defense.

What changed

• House Bill 184 (136th G.A.) added R.C. 9.66(D). It makes “any” information submitted to a political subdivision, port authority, or tax-incentive review council in connection with economic-development assistance confidential and not a public record under R.C. 149.43 — and exposes officials who disclose it to liability, including criminal penalty. (Policy Matters Ohio; Bricker DevelopOhio)

• How it became law matters. HB 184 was an omnibus vehicle — college-athlete NIL contract limits, appropriations, and amendments to numerous Revised Code sections — and §9.66 was carried in as a Senate amendment, the “little-noticed” addition the commentators describe, not a standalone economic-development bill. Provenance:

  • 2025-11-19 — House concurs in the Senate amendments (the vehicle that inserted §9.66 into the bill);
  • 2025-12-19 — Gov. DeWine signs HB 184 (enacted); appropriations sections take effect immediately;
  • 2026-03-20 — §9.66(D), being non-appropriations, takes effect on the 90-day date — the pivot for the §1.48 argument below.

  The one gap is the exact Senate Finance committee amendment that inserted §9.66 — the precise date and sponsor of the insertion — which the records campaign has not yet pinned (“our only missing record”). That the confidentiality regime entered law as an unattributed amendment to an NIL/appropriations omnibus is itself a transparency point worth making to the committee.

• This explains the county’s behavior far better than mere foot-dragging. “Forwarded to legal counsel, no schedule” is what a risk-averse office does when a brand-new statute threatens its own staff with criminal exposure for disclosure. The delay is likely §9.66(D) caution, not indifference — which is both an explanation and, for the county, a colorable (if overbroad) defense.

Why the cure notice’s counter is the right one

Relator pre-empts §9.66(D) on two textual axes, and both are well-aimed:

1. Temporal — R.C. 1.48 (statutes are prospective). Records predating March 20, 2026 are presumptively outside a statute that took effect that day. Much of the BOSC paper trail — the 2025 AEDG/board activity, the pre-March infrastructure and wastewater planning, the website CMS history the request targets — predates the effective date. This is the single best wedge: §9.66(D) cannot retroactively re-classify already-public records absent express legislative intent, which §1.48 presumes against.
2. Categorical — “submitted to… from an applicant.” §9.66(D) protects information the applicant submits to the subdivision for economic-development assistance. It is a far weaker fit for:
   • the county’s own work product (cost-benefit analyses, staff evaluations);
   • intergovernmental communications (county ↔ townships, JobsOhio, OEPA, DoD);
   • infrastructure/utility and NPDES/PTI records (the pump station RFP, Shawnee II Phase 2 capacity, the forcemain) — these are public-works and environmental records, not “economic-development assistance” submissions, even if the same project drives them;
   • the existence and text of NDAs themselves and the §9.66(D) withholding justification, which the requester is entitled to test via a withholding log.

The right ask — which the complaint makes — is therefore not “produce everything” but “produce a record-by-record withholding log identifying each §9.66(D) claim, the factual basis, and whether it pre- or post-dates March 20, 2026.” That converts a blanket “it’s all confidential now” into individually contestable determinations, and it is the procedural posture §9.66(D) is most vulnerable to.

The legislative tide is moving the other way

Two contemporaneous developments are worth putting in front of any court as context for why transparency here serves a live public interest:

• HB 695 (Reps. Brian Stewart & Adam Bird, ~Mar–May 2026) would prohibit certain local officials from signing economic-development NDAs and (per pro-transparency advocates) repeal the §9.66(D) addition, restoring public access to data-center deal records — an express legislative signal that §9.66(D) is contested and reform is in motion. (Policy Matters Ohio, May 20 2026)
• Gov. DeWine paused new data-center sales-tax exemptions (R.C. 122.175) pending a General Assembly study, after the program cost ~$1.6B against a $136M projection (a ~$1.4B 2025 overage). (The Hill; Signal Ohio)

§9.66(D) and the tax-exemption pause are the same policy story from opposite ends: the state simultaneously subsidizing hyperscale data centers and classifying the records that would let the public price those subsidies. The BOSC request sits exactly on that fault line.

The committee’s bill slate (where §9.66(D) sits in the larger debate)

§9.66(D) is the confidentiality counter-current inside a much larger data-center reckoning now before the Select / Joint Committee on Data Centers (record indexed at data/extracted/legal/select-committee-2026/select-committee-2026.hearing-index.yaml). The bills in motion:

• HB 706 (Reps. David Thomas & Tristan Rader) — the core consumer-protection vehicle: makes cost causation enforceable (long-term service agreements, minimum billing demand, exit fees, a bar on shifting data-center costs to other ratepayers) and extends the PUCO-confirmed AEP Ohio tariff statewide.
• HB 646 — a data-center study commission (the natural forum for the §9.66(D) repeal and disclosure questions; vulnerable to industry capture without independent facility-level data).
• HB 695 (Stewart & Bird) — the NDA-prohibition / disclosure-restoration bill above.
• Electricity Forecast Integrity Act (EFIA) — anti-speculation load forecasting.
• A proposed constitutional amendment (Ohio Prohibition of Construction of a Data Center Amendment, 2026) defining a data center by a > 25 MW peak/aggregate load — but with no ownership-disclosure or parcel-aggregation mechanism, which the testimony argues hyperscale routes around.

The throughline the relator’s testimony supplies: every one of these depends on knowing who the customer is, and §9.66(D) is the statute pulling the other way. That is the frame that makes a single resident’s records fight legislatively relevant — it is a live instance of the disclosure gap the whole slate is circling.

The hearing record undercuts §9.66(D)‘s own premise

§9.66(D) rests on an unstated assumption: that economic-development information of this class is categorically confidential, and that the confidentiality serves a public development interest strong enough to override R.C. 149.43 and even criminalize disclosure. The same committee’s own hearing record — the hyperscalers and developers testifying under oath-equivalent on 2026-06-04 — contradicts both halves of that premise in their own words. This is not advocacy gloss; it is the regulated industry, on the legislative record, describing what it actually treats as secret and why. (Transcripts and indices: data/extracted/legal/select-committee-2026/hearings-audio/bosc-committee-hearing-2026-06-04-{am,pm,pm2}.*; machine- transcribed on-device, timestamps are mm:ss and should be verified against the audio before quotation in a filing.)

1. The confidentiality is not industry-universal — some operators publish the very data §9.66(D) shields. Asked point-blank by Sen. Blackshear whether there was “a reason not to publicly disclose electric and water usage” (pm, ~62:00), the panel split:

• Microsoft (Daniel Brown): “No reason. Historically it has been proprietary information, especially on electricity usage, but Microsoft will report those in our sustainability reports yearly.”
• Google (Liz Schwab): “We report all of our operational electricity and water use in our annual sustainability report.”
• Amazon Web Services (Craig Sunderman): reports water “site-by-site” annually.

A statute that declares an entire category of records confidential as a matter of law is hard to defend when the category’s largest actors voluntarily publish the same figures. Where the legislature has classified information that the affected parties themselves treat as disclosable, the public-records presumption of openness (which R.C. 149.43 directs courts to construe liberally, with exemptions narrowly) should reassert itself.

2. Where confidentiality is asserted, the interest is ordinary commercial competition — and even the competitors who share it disagree that it requires secrecy. AWS, on the energy figures it does guard, was candid about why: that data is “fairly sensitive market-share information in terms of the cloud providers… you can directly tie it back to cloud market share, which … we compete on vigorously” (pm, ~63:26–63:37). Two things follow, and neither supports a blanket statutory exemption:

• The interest is specific to public cloud platforms — and is not even shared across them. The concern arises only because a public cloud operator’s energy draw can be reverse-engineered into the sellable compute capacity it markets by the unit. That describes AWS, Microsoft Azure, and Google Cloud — direct competitors in one market. Yet of those three, two disclaim the secrecy: Microsoft (“No reason… we will report those yearly”) and Google (“we report all of our operational electricity and water use”). The objection is therefore not an industry position, nor even a public-cloud- industry position — on this record it is one competitor’s preference, contradicted by its two closest rivals. Meta’s willingness to “aggregate” is not a fourth vote here: Meta is a captive, first-party operator — it runs data centers for its own products, not a public cloud it sells by the unit — so its figures carry no cloud-market-share signal, and its comfort with disclosure costs it nothing (its aggregation remark was about water, not the energy figure at issue). The point cuts the opposite way from the casual reading: §9.66(D) shields captive operators like Meta despite their having no market-share interest to protect at all.
• The genuine interest is already met by narrower tools than a public-records exemption. A competitive-position interest in a discrete figure is the province of trade-secret law (R.C. 1333.61; the R.C. 149.43 trade-secret line of cases), which shields the specific sensitive datum without classifying the whole category. And AWS itself named the proportionate mechanism: in “many jurisdictions… we report energy usage… whether that’s held confidentially at the regulator space or reported to a state agency” (pm, ~63:52) — i.e., confidential reporting to the regulator, not a blanket bar on public access to every economic-development record.

§9.66(D) is therefore substantially overbroad relative to the interest the industry actually asserts: to shield one public-cloud operator’s reluctance to reveal a single competitive data point — one its own direct competitors disclose, and that trade-secret law and confidential-to-regulator reporting already handle — the statute sweeps in captive operators with no such interest, the county’s own work product, intergovernmental communications, and the existence of the NDAs themselves.

3. The opacity §9.66(D) protects can defeat even the developer’s own knowledge of who benefits. The disclosure gap is not hypothetical. Vantage Data Centers — a colocation developer — testified (pm2, ~31:51) that its tenants own the compute (“enterprise, cloud, or AI customers”) and that, asked who captures the sales-and-use tax abatement, it could not say: “I am not a tax policy expert… I do not know whether we directly take advantage and then it’s passed through as part of the financing, or whether it’s taken advantage by the tenants themselves.” When even the subsidized party cannot trace the public subsidy to its ultimate beneficiary, a confidentiality statute that locks the public out of that same question does not protect a development interest — it forecloses accountability for one. By contrast, the Bitcoin miner MARA volunteered the converse (pm2, ~11:38): “the one question I didn’t see come up that I thought was very pertinent is who has access to this data? … we don’t have customers in Bitcoin” — a witness independently identifying who-is-the-customer as the pertinent unasked question.

Why this matters to the filing. None of this changes the temporal (§1.48) or categorical (“submitted… from an applicant”) wedges in the cure notice — those remain the load-bearing arguments. But the hearing record supplies a public-interest overlay a court can weigh in construing §9.66(D) narrowly and in assessing the equities of mandamus: the General Assembly’s own fact-finding body heard the regulated industry concede that the shielded information is (a) voluntarily disclosed by its largest members, (b) protected, when at all, by an ordinary commercial interest that aggregation satisfies, and (c) capable of hiding the beneficiary of a public subsidy from the developer itself. That is precisely the posture in which a reviewing court should resolve doubt toward disclosure, and in which the record-by-record withholding log (not blanket confidentiality) is the proportionate remedy.

Presenting this draft as evidence to the General Assembly

The most strategically significant use of this document may not be the courthouse at all — it is the committee room. Presenting the unfiled draft to the General Assembly (as testimony supporting HB 695’s repeal of §9.66(D), or before the Joint Data Center Committee studying the industry) is a higher-leverage move than filing, for a structural reason worth stating plainly:

You cannot litigate your way out of a confidentiality statute. Mandamus is an after-the-fact remedy: it compels production of records the law treats as public. §9.66(D) operates upstream of that — it reclassifies the records as not public in the first place, at the threshold, before any duty to produce attaches. A writ can reach delay; it cannot reach a statute that says “this was never a public record.” So the mandamus treats the symptom (this county, these records, this delay); the legislative venue treats the cause (the exemption itself). Bringing the draft to the General Assembly aims at the root the courtroom can only work around.

That reframes why holding the draft, rather than filing it, can be the stronger play:

1. It converts an abstract policy debate into a documented, local, dated constituent harm. Committees hear advocacy organizations argue §9.66(D) in the abstract (Policy Matters). A resident’s actual stonewalled records campaign — with a court-ready, properly-pleaded complaint behind it — is the concrete case study those arguments lack. It is the single most persuasive register of legislative testimony: this statute did this, to me, about this $500M subsidized project, on these dates.

2. The unfiled posture is a feature here, not a defect. As legislative evidence the draft is better unfiled:

   • Not sub judice. Nothing is pending before a court, so neither you nor any legislator risks commenting on active litigation; you can speak the record freely.
   • It demonstrates credible capacity and intent to litigate — a complete pleading with the mandatory (C)(2) affirmation — which shows §9.66(D) chills oversight even from a citizen prepared and able to sue. That is a sharper proof of chilling effect than a complaint that was simply filed and resolved.
   • It keeps the litigation and legislative tracks separable, so testimony does not hand the county admissions to use if you later file.

3. It supplies the factual findings a repeal bill needs. HB 695’s premise is precisely that §9.66(D) blocks data-center transparency. “Forwarded to legal counsel, no schedule, no log” — a county invoking confidentiality over a hyperscale project — is direct evidence for that premise. And the substantive payload is already in this repo: the records the county is shielding are the water-allocation, sewer-capacity, and subsidy terms that the BOSC hydrology analysis shows matter (a ~24× 7Q10 cooling draw; a wet-weather surcharge exceeding both plants’ headroom on a system under a 2005 SSO decree). That is exactly the water/energy/land impact the Joint Data Center Committee was convened to study — you would be handing it a worked example.

Cautions, so the testimony lands. Testify as a fact witness to §9.66(D)‘s operation — your lived experience and the documentary record — not as a lawyer arguing its construction; the former is credible and nearly unrebuttable, the latter invites a statutory-interpretation fight you do not need. Pre-empt the obvious rebuttal (“the system worked — you got your confirmation”) by anchoring on the acknowledgment-without- production gap and the §9.66(D) overhang, not on whether one citizen eventually pried documents loose. And do not overclaim the defense/national-security dimension (below) without evidence — speculation there would forfeit the credibility the documentary record earns you. Legislative speech is largely privileged, but coordinate the two tracks deliberately rather than letting one contaminate the other.

III. Broader Ohio data-center development

The opacity pattern this case documents

The BOSC corpus is a clean specimen of how Ohio hyperscale projects are structured to minimize public visibility, and §9.66(D) now ratifies that structure in statute:

• Code names and shell entities. “Project BOSC” / “Project Dazzler”; Bistrozzi Addition LLC and Tilted Gate LLC (Delaware foreign LLCs registered through CT Corporation), with Google confirmed only on March 16, 2026 via AEDG — three weeks before Bistrozzi Addition LLC was even registered (Apr 6). The request (Ex. B) names the mechanism precisely: Vorys, Sater, Seymour & Pease and Scott J. Ziance, who leads that firm’s Ohio tax-incentive/CRA practice.
• NDAs as default. The original request (Ex. A, Category 1) specifically seeks NDA records and argues “NDAs executed in connection with public economic development transactions do not categorically exempt underlying public records.” §9.66(D) is, in effect, the legislature converting that private NDA practice into a public-records exemption.

The Google Lima project, confirmed

Public reporting now corroborates the corpus end-to-end: a ~$500M Google data center on 200+ acres in American and Sugar Creek townships, Google’s third in Ohio, construction 2026→2030, ~50 FT jobs, $250k/yr to Elida schools, a 15-year tax abatement, $50M of infrastructure (incl. the Bluelick/Cole and SR-115/Cole roundabouts — the very Tetra Tech OPC estimates extracted in this repo) and $13.6M in water infrastructure. OEPA is reviewing 115 emergency gensets, 36 cooling towers, and 113 fuel-storage tanks. (DatacenterDynamics; LimaOhio.com)

Those numbers are not incidental to the records fight — they are why it matters. The genset/cooling/fuel inventory is the same data the BOSC hydrology work uses to derive a 3.1–10 MGD consumptive cooling demand (~24× the Ottawa River 7Q10) and a wet-weather sanitary surcharge that exceeds both receiving plants’ documented headroom on a collection system already under a 2005 SSO-elimination mandate. The public interest in the withheld records is concrete: water allocation, sewer capacity, and who pays for the $50M+ of infrastructure are precisely the “economic-development assistance” terms §9.66(D) now lets the county decline to show.

Where this case fits statewide

This is one resident’s mandamus, but it is a microcosm of the live Ohio fight: HB 184 (§9.66(D)) vs HB 695, the $1.6B exemption overage and DeWine’s pause, and a General Assembly study committee weighing data centers’ water, energy, and land impacts. The mandamus mechanism (R.C. 149.43(C)(1)(b)) is one of the few tools a private citizen still has to force the issue record by record — and §9.66(D) is the statute purpose-built to blunt it. The temporal wedge (§1.48, pre-March-20 records) and the categorical wedge (public-works/environmental records ≠ “assistance” submissions) are the two openings, and this filing identifies both.

IV. Merged timeline — §9.66(D), BOSC, and the defense nexus

Three strands, one chronology. The pivot is March 20, 2026: everything above it is a public record by default; everything the county now shields under §9.66(D) must postdate it (the §1.48 wedge). Tags: [LAW] statute/legislature · [BOSC] project + records campaign · [DEF] defense/DoD nexus. Confidence: [v] verified (corpus/web/filing) · [i] inference/approximate · [?] event known to have occurred, exact date unknown.

Many [BOSC] resolution events below are now corroborated at the page level against the Commissioners’ own published minutes — data/extracted/commissioners/minutes/ (634 meetings, 28 BOSC / 8 Bistrozzi page-hits). E.g. Res #417-25 (NDA) M052725.pdf:3, #304-25 (CRA) M041725.pdf:1,5, #548-25 (CRA) M071025.pdf:4; the deferred forcemain resolutions #113-26/#136-26 sit at M021926.pdf:3/M022626.pdf:3.

On the defense strand, stated honestly. Two things are verified and entity-level: Google Public Sector holds DoD authorizations up to IL5/IL6 and Top Secret and is a JWCC vendor; and the JSMC (Lima Army Tank Plant) is an active DoD armor facility in the same American/Shawnee Township corridor the PRR targets (Category 5). What is not established is any site-level link between the two — that the BOSC Lima data center will host government/defense workloads, or that JSMC adjacency is causal. The corpus shows only that the request asks about it (Category 1 “Google Public Sector,” Category 4 “U.S. Department of Defense entity”). Treat the defense nexus as an open question the records would answer, not a finding — and, per §II, do not advance it as fact in testimony.

Bottom line

• As a filing: procedurally clean (the (C)(2) affirmation alone puts it ahead of most pro se actions), correctly routed, ripe on the merits, with a fixed-at-cap $1,000 damages claim + costs. Date reconciled to April 14 (service date per the Ex. A header); still pin the request universe (the county’s “four” vs the pleaded three).
• As a legal matter: the merits of delay = constructive denial are strong; the contest is §9.66(D), and the winning move is the withholding-log demand plus the pre-March-20-2026 / non-”assistance” carve-outs, not a blanket production order.
• As public-interest research: the case is a precise instance of Ohio’s data-center transparency problem — code-named shells, NDA-by-default, a fresh confidentiality statute, and behind it the water/sewer/energy stakes the BOSC hydrology analysis quantifies.

Sources: R.C. 149.43 and 9.66 (Ohio Laws); HB 184 / HB 695 and §9.66(D) (Policy Matters Ohio, Bricker DevelopOhio); data-center tax-exemption pause (The Hill); Google Lima project (DatacenterDynamics, LimaOhio.com); Henderson v. Washington Court House, 2026-Ohio-110. Full case texts not independently verified. Primary corpus for the [BOSC] resolution strand: data/extracted/commissioners/minutes/ (Allen County Commissioners agendas + minutes, 2024–2026).