GCFT-0660 - Prospective Residual Tests of Relativistic Clock Transport with ACES: Synthetic Identifiability and External Wettzell Cross-Validation

This work presents a prospective, precommitted residual-test architecture for relativistic clock-transfer measurements with the Atomic Clock Ensemble in Space (ACES), combining the primary scientific analysis with its technical validation supplement. A forward-modelled synthetic measurement twin evaluates the frozen R3 statistic under an ordinary H0 and a phase-uniform 10−1510^{-15} injected H1 signal class. Across 10,000 realizations per branch, the frozen analysis yields a 1.00% H0 false-positive rate and 70.74% H1 detection efficiency. A read-only external H0 cross-validation using reconstructed Wettzell-like cadence yields a 0.77% false-positive rate and broadly compatible pass-count, duration, and gap statistics, while revealing a material orbital-phase-coverage discrepancy that worsens R3 conditioning from 16.706 to 43.471. Sensitivity analysis identifies correction-state fidelity as the dominant methodological limitation. The package includes deterministic replay, provenance controls, branch-isolation receipts, external-validation matrices, and reproducibility artifacts. No authoritative ACES flight rows are analyzed, and no observational GCFT detection or parameter estimate is claimed.