It is normal to get variation in those measurements, but it takes a little finesse to narrow it down.
The good news is that those measurements are not that sensitive and you get away with a little uncertainty in those values.
With the OAL Gage, the bullet wedges into lands based on friction and contact stress. There is always some variation on that depth just due to the shallow angles and materials.
With the comparator tools, there can be some error in the parallelism of the clamped on tooling, plus the difficulty of manipulating the sample to sit straight.
Some of this is due to the jaws of the calipers being offset to each other. There is an anvil that makes that a little less sensitive to technique, but even that tool cannot sit dead flat in line on the calipers.
There are other ways to make a sample cartridge at the length to touch lands, and all of them have their pros and cons. Even then, if you use caliper mounted comparators, you will need some finesse.
Some other comparator tools exist, but the ones that are dedicated to OAL or CBTO measurements are more expensive than the ones you clamp onto your calipers.
You may or may not want to put more effort or resources into this, but to be honest the whole goal is to get a starting point (CBTO at touch) from which to experiment with seating depth. Almost any number you start with good or bad, will need to be tested and the absolute value won't matter as long as you know how to repeat your setting. It will be your target tests that establish a hard number.
