CIP: normalize final evidence on rift_O4c

[oshaughnessy-junior]

Summary

Port the final-CIP evidence normalization from #177 onto the older rift_O4c release line.

• schedule strict evidence consolidation for the final fixed CIP iteration
• run an independent terminal CIP integral with L=1 using the configured prior and corrections
• report ln B_H = ln Z_H - ln integral(prior), including propagated Monte Carlo uncertainty
• preserve the historical two-column evidence output
• serialize final evidence before convergence can terminate the DAG

O4c adaptation

• retain the O4c pipeline implementation and do not resurrect the legacy cepp_basic_htcondor driver absent from this branch
• add the evidence consolidator, which did not yet exist on O4c
• avoid O4d-only supplementary-likelihood offset and hyperpipeline helpers

Validation

• 6 focused consolidation and pipeline tests pass
• py_compile passes for all three modified executables
• generated an exploded two-iteration O4c DAG with CIP_prior and evidence_final submit files
• verified emitted ordering: final CIP product -> prior integral -> normalized evidence -> convergence test
• completed a direct O4c L=1 CIP smoke run with finite normalization, neff, and Monte Carlo error
• private RIFT review dispatch passed at f29a0c2

The O4c branch does not schedule the newer full CI workflow used by rift_O4d; its available private review check and all branch-specific local gates passed.

[interhost-cross-review-canary-v2]

Independent automated review completed at the recorded exact commit. Detailed findings were withheld from public output by the private-context egress policy and require private human declassification.