Completed clustered-boundary comparison
Shape result: globe. All six primary fits passed: True. Both controls passed: True. Required repeat agreement passed: True.
The completed diagnostic used 54 of its 58 permitted native calls; this review performed no additional fits or predictions. The existing timing search located a stable point within a cluster of closely spaced integration boundaries. Both sides passed the unchanged final gate. Five companion primary fits were verified from sealed artifacts. The new diagnostic repeat starts from a separately verified nuisance solution; it does not repair or relabel the interrupted free-time fit. Earlier failures, measurements, bounds, clock and receiver weights are preserved.
| Repeat check | Measured difference | Required limit |
|---|---|---|
| Relative cost | 1.34632e-11 | 10⁻⁶ |
| North track RMS | 1.33448e-06 m | 0.01 m |
| East track RMS | 2.4616e-07 m | 0.01 m |
| Up track RMS | 3.84778e-08 m | 0.01 m |
| Processing case | Globe with field cost | Zero-field globe cost | Flat-disc cost |
|---|---|---|---|
| bias1_profiled_removal | 4.362657 | 2570.949968 | 8437.588092 |
| bias1_unsubtracted | 4.362657 | 2570.936955 | 22203.379091 |
These are numerical model preferences and repeatability checks, not calibrated significance or instrument-accuracy estimates. Background-field preference follows the shape comparison; the field’s cause and solar-versus-sidereal rate are not determined.