CRYOPULSE P4.4 / REVIEW COMPANION
Physical measurements: 0. No performance score is changed by this companion.

Use alongside the immutable vehicle source archive:
CryoPulse-P4.4-model-and-results.zip
SHA256 70a744e6d33821328844f6d2cf7162cca866eb750329477b366bb48ec18513f5

Extract both packages into the same parent folder. The companion adds files;
it does not replace the core FILE-SHA256.json or alter the frozen vehicle model.
Its separate FILE-SHA256-review-companion.json verifies every added file.

1. Fresh-result replay
The staged verifier deletes archived selected results from two clean extractions,
recomputes the primary and reserve launches independently, then checks all four
speed policies. Read cryopulse-p44/site-source/REPLAY.txt. The supplied replay
report has zero differences in the reported performance quantities. The former
single-session run reached its execution limit; bounded stages completed.

2. Bench precision planning
The unchanged twelve-pair CF-1.1 decision is evaluated with a stated independent
normal paired-log-ratio model, its planned calibration bounds and a conservative
head-mismatch penalty. The noncentral t calculation is checked against 200,000
virtual repetitions per condition. This estimates operating characteristics
under those assumptions; it creates no instrument records or physical evidence.
The effect cases come from the independent 1D input-sensitivity study and are
not a confidence interval for the real effect. Run:
  python cryopulse-p44/scripts/bench_power_study.py

3. Tyre requirements
The selected event's saved states become a load-versus-speed supplier brief.
It does not qualify an existing tyre, provide a rigorous inter-sample peak bound
or resolve the separate top-speed event's full load/stability envelope. Run:
  python cryopulse-p44/scripts/tyre_requirements_brief.py

4. Assembly and heat measurement
The bench review notes distinguish gap/clamp control, actual pressure planes,
optional gas-side calorimetry and the complete electrical/thermal event. The
primary mass-flow endpoint, twelve-pair count and frozen controls stay unchanged.

The figures are computed scientific plots. The data are calculations, not road
or laboratory measurements. The whole vehicle still needs digital and supplier
qualification as stated in the current audit and website.
