The most expensive data infrastructure is the infrastructure built after the disaster

The conclusion of the paper, in one sentence: the most expensive data infrastructure is the infrastructure built after the disaster.

The same architectural decision — deploying Tier 3 preservation on existing institutional infrastructure — costs effectively zero today and addresses the latent §5 liability before the §5.4 trajectory surfaces it. Built after the disaster, it costs:

• The settlement value of any FCA case that lands during the gap (E-0041, E-0042, E-0043 establish the $10M-$112.5M settlement range).
• The forensic-recovery effort required to reconstruct what can be reconstructed (Term B in M-0003).
• The downstream-research value never realized (Term C).
• The faculty-flight and failed-recruiting cost compounded across the years of underinvestment (C-0026).
• The competitive position lost to institutions that deployed the architecture earlier (C-0008's two-sided position cedes to those institutions).

The 1:10:100 cost heuristic (M-0004) holds at minimum. The asymmetry between prevention and post-disaster reconstruction in this specific case may run closer to 1:1,000,000 — the gap between Tier 3 deployment cost (effectively zero on existing infrastructure) and the latent liability (~$1.1B/year at a $200M R1, C-0005).

The recommendation regime developed in §11 (R1-R7) is the operational expression of this asymmetry. Each recommendation has a specific audience and timeline. None of them require new technology to be invented or new architectural principles to be developed. They require the institutional decision to deploy what already exists, before the moment when it is required to have already been deployed.

This is the closing Claim of the paper. The architecture that produces these outcomes is already deployed at scale in adjacent domains. The remaining requirement is deliberate research-sector adoption.