Our FR5 planning sheet lists how far each hand-loaded blank sticks out but not how long the blank is, and somehow the conclusion says inconsistent seating, which seems a remarkable amount of certainty from half a measurement.
Pair overall blank length with exposed length for each identified sample. Also agree the chuck reference with the setter. Otherwise two people can make equally careful measurements of different things.
Have the setter define the intended seating and manual comparison method before taking another set of readings, while automatic loading remains outside this check.
Setter and inspection agreed the reference and marked each blank, and the first paired measurements explain most of the spread through actual stock length, which the spreadsheet had politely ignored
One blank gives a larger exposed reading on some repeats while ordinary ones are consistent, and the setter says that one feels different at the stop, so it is retained separately with its readings
Keep the blank's original condition documented and let the setter and inspection assess how that edge meets the intended stop, rather than changing the loading allowance around it.
They confirmed the raised edge makes that blank catch before full seating; ordinary conforming blanks remained repeatable, so the estimate now separates normal length variation from this rejected blank's seating fault