Old bracket fixture has a clamp nobody can explain

Designer marked it as restraint only. The old inspection sheet measures the upright while clamped. The finished drawing specifies the released assembly. Those are different conditions.

Then preserve those conditions in the review, rather than comparing the measurements as though they were equivalent. Do any retained parts have incoming fit-up measurements linked to their final inspection records?

That would be useful context for the fit-up question. An apparently unexplained correction can become understandable once the component history is visible, though it still needs assessment for the new process.

Only the bracket number is consistent. Two retained samples have different gusset revisions. Quality is separating them before measurement; we cannot attribute the old spread to the clamp alone.

That is a defensible limitation. Put ownership of the new fit-up and final acceptance checks into the requirements so the next fixture review has evidence the old records lack.