Theory of Everything Research Programme: TP2-D1 Fixed-Horizon Heritable Fitness Coupling — Confirmatory Dataset and Decision Record.

Confirmatory dataset and decision record for the preregistered TP2-D1 experiment in the Theory of Everything Research Programme. TP2-D1 tests whether parent-specific inherited structural state predicts differential descendant production over a fixed three-generation lineage horizon and whether that coupling produces directional population change. The study was preregistered on OSF at DOI 10.17605/OSF.IO/9FMYA. The experiment is a new follow-up to the original TP2 confirmatory study. The original TP2 result remains TP2 NOT RETAINED under its direct-daughter fitness criterion. TP2-D1 prospectively tests a different multigeneration lineage-fitness estimand rather than reinterpreting the failed original endpoint. The confirmatory programme executed all 40 registered candidate populations together with all paired causal intervention and control branches. There were no replacement seeds and no technical missing runs. The primary results were: HTP2-D1.1 — Fixed-horizon fitness variation: 40/40 PASS. HTP2-D1.2 — Genealogy-specific heritable fitness coupling: 40/40 PASS. HTP2-D1.3 — Causal inherited-fitness coupling: 39/40 PASS. HTP2-D1.4 — Parent-shuffled genealogy rejection: 0/40 bypasses, PASS. HTP2-D1.5 — Fitness-decoupling rejection: 0/40 bypasses, PASS. The candidate populations showed a median of seven informative lineage depths and median selection-response alignment of approximately 0.857. Genealogy-permutation excess was positive in all 40 candidate populations. The causal intervention branch showed: knockout covariance below baseline in 40/40 populations; rescue covariance above knockout in 40/40; rescue closer to baseline than knockout in 39/40; iso-effect substitution within 20% of baseline in 40/40. The matched alternative-architecture branches remained inconclusive at the architecture level: HTP2-D1.6 — non-maintaining architecture: 4/40 adequately matched, with 4/4 matched controls satisfying the full bypass profile. HTP2-D1.7 — heuristic architecture: 6/40 adequately matched, with 6/6 matched controls satisfying the full bypass profile. Because fewer than 20 controls were adequately matched in either branch, both were classified MATCHABILITY-LIMITED / INCONCLUSIVE under the preregistered rules. The final preregistered decision was: TP2-D1 QUALIFIED SUPPORT — ARCHITECTURAL SPECIFICITY UNRESOLVED. The confirmatory evidence therefore supports, within the registered artificial lineage domain: heritable variation → differential fixed-horizon lineage fitness → directional population change. It does not establish architectural uniqueness or necessity. This dataset contains the complete population-level confirmatory results, hypothesis decisions, diagnostics, execution verification, final unlock record and machine-readable decision record.