The installed source passes all native controls

All 24/24 controls passed in 201.31 seconds. The study charged 78 full native evaluations, 348 complete bounds and 30,144 world/source bound steps. Source-reference and rounded-body work were reserved separately.

The revised proof uses the actual asymmetric calibration box. It retains the original physical bounds and nonlinear caps, reference centers, all 51 columns and every acceptance limit. The six broad cases that stopped the first candidate now pass. That earlier failed study remains preserved.

This qualifies the integrated candidate on the original generated control matrix. It does not establish useful complete prediction widths or calibrated observational coverage. No aircraft fitting occurred.

The six-source comparison now tests the new allowances against the best saved compatible bounds. Complete sector unions and the four-of-six width target remain a later gate.

Evidence · Fixed scope · Next task.