I decided to illustrate the principle with exaggeration so the effect of ogive variation can be seen. You can see from the illustration that a smaller-than-throat hole contacts the different size ogives at different heights. This is true whether the smaller hole is in a seating stem ram as shown, or if it is a comparator insert. An actual throat profile, though, will contact the bullet where it actually meets the lands and this eliminates and ogive variation causing an error in the measurement.
The practical problem stems (cough! cough!) from the seating ram being in the same location from the case when seating both bullets. It results in ogive error at the seating stem's contact point being added to or subtracted from bullet jump, depending on which direction the ogive irregularity was in. You can see the bullet bases and throat contact points being offset as a result of this.
Then there is the problem of uneven case shoulder difference in resizing on rimless bottleneck cartridges. Since the case shoulders stop the case going forward, if one shoulder is a couple of thousandths shorter than the next, even if the bullets were absolutely identical, the bullet loaded into that case will have that couple of thousandths less jump, since it sticks that much further into the throat when the case shoulder hits the chamber shoulder. Below the bullets are identical and the case lengths identical and they are seated with the same seating die setup, yet, because the second case's shoulder "setback" is greater, the bullet jump upon firing is less.
The only way I know to KNOW your bullet jump is the same is to measure the case shoulder location and the bullet's actual contact point with the chamber throat and make sure the comparative difference is the same.