Declared before new diagnostic calculations. This is a descriptive interrogation of already recorded experience, not a new collection or fit. Evidence identities and immutable parent pins come from each native reader; do not infer admission from a file existing.
read_generation_posthoc_report
and the separately archived factual-prefix summary. Primary question:
does observed late-prefix support loss justify claiming H16 unnecessary?
Report reached/intended chains, scene support and paired late gains on
their own risk sets. Do not subtract medians from different completer
populations. Historical fixed-root cumulative gain is a diagnostic
proxy, not an independently measured endpoint. No causal stopping-length
effect, new horizon choice or cost-optimal cutoff is estimable from
these aggregates. If exact aligned per-prefix clocks and independent
endpoint outcomes are absent, report that gap rather than inventing
marginal quality per second.Relocated on 16 October without a new preregistration claim. Complete original declared protocol.