Complete common-force-transform prediction comparison
48/48 sectors pass correctness; state: complete.
All six cases meet the preregistered twofold width-reduction target: no.
Complete generated prediction widths at 7.2 seconds; all original physical boxes, eight sectors and 51 columns remain. These are not measured aircraft errors, calibrated intervals or new model preferences.
| Candidate / removal | Complete | Width N / E / height (m) | Ratio to projected baseline | Twofold target |
|---|---|---|---|---|
| Globe with specified field / zero | Yes | 1.2729, 1.6489, 1.0761 | 0.5064, 0.5861, 0.5846 | No |
| Globe with specified field / profiled | Yes | 1.3584, 1.7238, 1.1626 | 0.5226, 0.5967, 0.6034 | No |
| Globe with zero field / zero | Yes | 0.9782, 1.0536, 0.7584 | 0.4334, 0.4558, 0.4694 | Yes |
| Globe with zero field / profiled | Yes | 1.1344, 1.1995, 0.9145 | 0.4700, 0.4854, 0.5158 | No |
| Flat disc / zero | Yes | 0.7005, 0.2203, 0.5252 | 0.3462, 0.3577, 0.3434 | Yes |
| Flat disc / profiled | Yes | 0.8894, 0.2791, 0.7080 | 0.4031, 0.4160, 0.4132 | Yes |
The study charged 144 native calls and 384 bounds in 2985.6 seconds. Bound-step reservations count both exact-clock and propagation passes. No original aircraft reads, observed fits or protected ESA outcomes occur. Recording-scale usefulness, whole-program numerical treatment and observation-error adequacy remain separate gates.
Frozen scope · Joint qualification · Projected baseline · Roadmap.