Common acceleration: complete comparison

All 48 sectors pass correctness. All six cases also pass the unchanged tenfold width-reduction target against the original broad-box baseline. Complete union widths range from 0.061 to 0.358 m across axes and cases.

Candidate / processing	Complete	North / east / height union width (m)	Ratio to original broad-box widths
Globe + specified field, zero removal	Yes	0.277, 0.350, 0.194	0.000361, 0.000596, 0.000378
Globe + specified field, profiled removal	Yes	0.286, 0.358, 0.202	0.000371, 0.000608, 0.000394
Globe, zero field, zero removal	Yes	0.225, 0.236, 0.138	0.000293, 0.000402, 0.000269
Globe, zero field, profiled removal	Yes	0.243, 0.253, 0.156	0.000316, 0.00043, 0.000304
Flat disc, zero removal	Yes	0.183, 0.061, 0.123	0.000238, 0.000354, 0.000241
Flat disc, profiled removal	Yes	0.205, 0.068, 0.145	0.000266, 0.000391, 0.000284

These are conservative calculated ranges at 1.8 seconds in generated data, not measured aircraft errors, calibrated prediction intervals or a new model preference. All original physical models, 27 calibration directions and bounds remain. Fifteen auxiliary coordinates preserve common errors across measurement times; they are not new independent physical choices at each epoch.

The study charged 144 native evaluations and completed in 1432.6 seconds. Source/input/environment freezes, native values/Jacobians, training inequalities and future support certificates remain available. This report verifies the published hashes and reconciles every complete union against its sectors, without numerical integration, optimization or original-file reads.

Longer recording-scale usefulness, whole-program floating-point treatment, actual receiver-error adequacy and independent confirmation remain separate gates. No protected ESA outcomes are used. The 53-stretch empirical comparison is unchanged; scientific decisions remain abstention.

All saved results · Frozen method · Native qualification · Roadmap.