Correction for best fit rotation

[[Ja...]]

We have a part. OD is a circle. ID is an ellipse both on the same center.
There are two circles to the left and right to the centerline of the ellipse.
The top of the part is Datum A.
Left hole is Datum B and right hole is Datum C

ID is called out for Profile to ABC and Profile to A
Profile to A is showing us the translation of X and Y and the Rotation of Z.

We are wondering, is the rotation of Z from the center of the Ellipse feature or is it in relation to ABC?

[[Da...]]

Does the call-out have MMB?
on the Datum -B(M)- or Datum -C(M)- ?

[[Da...]]

As I am rereading your post I thing you are talking about have a Composite call-out

So the top profile is like .003 to ABC
and the bottom one is like .001 to A?

If this is the case then it will Translate in X Y and rotate about Z to fit the shape best. In this call-out A will only control the perpendicularity of the profile and not the X Y position nor the Z rotation

[[Ja...]]

So here is an image of something similar to what we are trying to do.
We are reporting the profiles as shown.
What we are trying to figure out is, where is the rotation taking place, so we can correct the orientation of the ID to the datums.
Say the ID is rotated 0.5° and we want to correct it. Currently, we are using the rotation value that is given on the Profile to A report (also attached, but shows a good part). But the engineer says he doesn't believe it correlates with the adjustment to the machine. Just trying to understand what the report is telling us so we can correct it.

Capture.PNG