Our bracket passed a clamped check that never measured the failed dimension

Hole position relative to the base, according to the instruction inspection found. Not upright height. So the released height failure can be real without showing how far it moved from a measured starting height.

Have inspection confirm the released height measurement against its own drawing requirement, since discovering that the pin checked something else does not make the later reading unreliable by itself.

And does the correction note identify which feature was altered before the final pass?

It names the upright, Sam, but not the actual correction method. Isabel, inspection confirmed the height check used the correct revision and reference face. I've stopped calling the pin check a starting-height reading.

Then the useful next comparison needs defined incoming fit-up and released measurements on identified brackets, with fabrication explaining the correction operation rather than asking the fixture proposal to make it disappear.

Would that also let the trainer show why passing one feature does not settle a different one?