A seating die stem contacts the ogive well out toward the meplat of the bullet. A caliper insert contacts the ogive slightly above the bearing surface close to where the bullet will first contact the lands. The bullet boattail and bearing surface are both well below these two points. Thus, sorting bullets by BTO dimensions is generally going to have little, if any, effect on seating depth as measured using calipers. In other words, the entire BTO dimension is outside the two critical contact points; i.e. the point near the meplat at which the seating die stem contacts and pushes the bullet down into the case neck during seating, and the point just above the top of the bearing surface where we actually measure CBTO (i.e. effectively, seating depth). Sorting bullets using the BTO dimension, which lies completely outside these two critical points is not going to affect seating depth.
What variance in bullet BTO can affect are the length of bullet bearing surface inside/contacting the case neck wall, and the effective/usable case volume. The first of these would relate to possible frictional differences between the bullet bearing surface and the case neck wall for bullets with different BTO dimensions, the second would relate to possible changes in pressure for a given load. The real question is whether the difference between either of these two variables would be large enough for bullets that vary in BTO dimension by only a few thousandths that it would make a significant difference on the target. Of course, that can only be decided by the individual after rigorous testing.
But consider what we are actually doing when we carry out a seating depth test. Effectively, a seating depth test is also moving the bullet boattail/bearing surface in/out of the neck, thereby potentially changing both the amount of bearing surface contacting the case neck wall, as well as the effective case volume. With regard to both friction between the bullet bearing surface/neck wall and the relative change in effective case volume, how much change in velocity/pressure does one typically observe during seating depth testing? The answer largely depends on the total test range over which the bullets are seated. Seating a jumped bullet .050" or further off the lands can noticeably increase pressure; changing the seating depth of a jumped bullet by only .010" or .020", not so much. If I cannot detect a significant change in velocity/pressure across a typical seating depth test with bullets seated from about .003" off the lands, to about .024" - .027" off the lands (essentially all of the bullets I shoot will tune in somewhere within this range), then as an F-TR shooter I am going to be relatively unconcerned about BTO variance of a few thousandths. For that reason, I also sort bullets by OAL only. Nonetheless, for shooting disciplines where the ultimate precision of a load is everything (i.e. BR), the effect of BTO length variance is a testable commodity. If you think it might make a difference, the easiest approach initially is to load a few bullets from the two BTO extremes (i.e. longest/shortest) and see if you can shoot the difference in terms of precision, or detect any difference in velocity/pressure. Then you will know whether sorting bullets by BTO will have a significant effect on precision in your hands.