The runtime obstacle is resolved on this screen

All six full-history source comparisons complete and preserve their reference maps. The study took 134.46 seconds; each bound finished within its original 120-second limit. The previous implementation completed only two cases before four globe cases reached that limit.

Compared with the best compatible saved bounds, the new height allowances are 17–37% smaller. North and east allowances are wider in every case. The retained comparison therefore keeps the old horizontal bounds and the new height bounds. It never substitutes a wider allowance for a tighter valid one.

This screen does not recompute complete prediction regions. The last complete width comparison remains four of six groups, and withheld aircraft fitting is still waiting for a useful complete error range.

The next task is to retain the source's calibration derivatives when bounding its nonlinear remainder. The current value-interval calculation loses that linear information and overstates horizontal uncertainty. A replacement needs its own mathematical and numerical qualification, including reference rounding and packet-clock boundaries; acceptance limits stay fixed.

Six-case results · Frozen scope · Technical next step · Roadmap.