Rejected exploratory joint comparison

Both methods used 0.03 m voxels, three local-map scans and a 1 nm coordinate policy; returned scan motions retained floating arithmetic. The complete six-second joint comparison failed the fixed quality limits. These working-tree explorations are not formal speed comparisons and their timing is not pooled with the formal cohort. Source revision fields alone do not identify these uncommitted changes. The exact bytes of every actually imported Calibrex module were reconstructed from the recorded hashes and retained here, with their manifests and the original helper. Only imported modules are archived for this exploratory pair, rather than claiming a complete frozen working tree.

The calibration artifacts validate with the ordinary joint-calibration schema. The private measurement records validate with the local measurement schema. The archive audit recomputes the rejection, helper/import/result bindings and per-round records. No quality tolerance was changed to accept this pair.