recalling points from 2D curve

[[Pa...]]

I'm scanning a 2D curve for a line profile call-out and then recalling a few of those line points as individual points and using those space point distances for a surface profile calculation. This give inconsistent results. The 2D curve is right where it is supposed to be with good form, the recalled points/space-point-distances are sometimes .010" off nominal--not what the line profile shows, but other times right where I expected them to be. I have no idea why. A few notes:
1) I am being consistent with my references to datums and not comparing an apples measurements (a line with with no datum reference) to SPDs that, of course, must work back to an alignment.
2) Measuring corresponding individual "real" points does give an answer consistent with the line profile. The problem is in the recall feature points function.
3) Recalculating the recalled points using actuals to nominals function does "fix" the problem, but is not appropriate to use: I'm then changing my nominal to match my actual so of course SPD reads zero.
Any ideas?
PaulS

[[Kl...]]

Hello,
hope the example helps you
Inspection runs in Simulation
pdf shows report

PointDistance.zip