# Additional read-only10us cut-recovery diagnostic

Registered while the new native time pair is running, before these new
cut-recovery observations. It does not amend the native frozen plan, source,
parameters or original acceptance gates.

At the10us saved step, apply unchanged prepared_motion/analyze.py:physical_cut
to both new uniform10cases and both frozen uniform20references. Independent
SI thermodynamic pressure, least-squares pressure gradient, lagged h^3/(12mu)
mobility and exact GEOS cut lengths recover a Darcy normal-flux estimate.
Recorded node velocities recover a separate material/monolayer cut estimate.
Compare all three components to their existing discrete ledger rates.

Preserve raw dimensional values as well as single-step relative differences;
do not call one sample a temporal L2 error. Native ledger denominators can be
small. This reconstruction is nonconforming and not independent Stokes flux
or continuum traction validation. Earlier kinematics identities do not turn
these reconstructed velocities into a proved physical flux. No flux is defined
as the missing conservation residual; no closure/kinetics/rate is installed.

This is an extra diagnostic of how cut recovery changes with actual whole
domain resolution, alongside (never replacing) the original dh/motion gates.
Full physical normal flux/traction and50/200us qualification remain required.