It says aligned, no figure or method. The released sheet has measured values. I was giving the first entry more weight than it deserves

Keep it, but describe it as an observation. Then ask for the part identity, fixture revision and measuring supports behind the released values. You need to know whether the sheets even describe the same configuration before interpreting the difference.

Are the released values before or after the correction?

And do not call that difference distortion yet. One side of the comparison has no measurement.

The dated sheets include both, Oscar, for one assembly. The correction note sits between them. I can connect that pair by identity, but not to any incoming measurements

Then take that one assembly as a worked records example. Ask the process owner to explain the correction and preparation stages. Leave the absent incoming values empty; an accepted final figure cannot supply them backwards.

Would the same inspection support be used in the proposed cell's final acceptance? If not, quality should resolve the method comparison before those old figures become its baseline.

I asked quality. The old sheet specifies a support arrangement that isn't in the proposal at all. They are reviewing the final inspection method now, separately from the nest's holding contacts

That is a concrete missing requirement. It still doesn't tell us why the bracket missed its dimension before correction.

Has the process owner explained the correction entry yet?

Not yet. Quality has added the final measurement support requirement to the proposal, but the correction explanation remains with fabrication. I have kept the two actions separate so the first doesn't accidentally close the second

That is where I'd leave the historical example for now. The forthcoming study needs to capture what this record cannot, rather than make the old pages sound more complete.