Lynn,
If you have the resources readily available, do me, and possibly yourself a favor and try the following.
Make an insert for your caliper ogive measuring attachment that hits the nose of your bullets at the same diameter as the seating stem does. Use it to measure (to their bases) a group of bullets that are either numbered, or in numbered positions. Record the measurement for each bullet. Next, substitute an insert that hits the bullet where the rifling will make contact. Measure the same bullets and record the measurments for each. Finally, subtract the second measurement from the first for each bullet, and tell us your results.
In effect, this is what Bob Greens tool does without the subtraction. I would be very interested in your results. If the distance between the circles varries then, it would seem to me that relation of the ogive to the rifling would, at a minimum varry by that amount, if neck friction was identical. Given that it is not, the variance that it (neck friction) introduces can either add to or subtract from the ogive variance. In a large sample, I think that it would increase the range.