When using the Hornady OAL gauge to determine the distance to touching, you are actually basing the measurement when the shoulder of the OAL gauge case is in solid contact with the chamber shoulder. A much easier way of stating that is simply that the OAL gauge case is held solidly in the chamber against the shoulder as the bullet is pushed out with the plastic rod to reach just "touching the lands. The actual measurement is then taken using calipers once the tool is removed from the chamber. That means you're measuring from the base of the case (i.e. NOT the bolt face) to a point on the bullet that is presumably close to where it first touches the lands.
All this is really does is to establish a reproducible measurement that can be used as a reference point for subsequent CBTO measurements during the reloading process...setting a kind of baseline measurement, if you will. It's a reference, nothing more. However, it's important to note that it's unlikely to be off by a substantial margin from the true value. Unless someone's technique was really very poor, it highly likely that this reference value might be within about .005" of the actual value. The same may not be true of a "ballpark" estimate, which could be off by .010", .020", or who even knows how much. So it's a "reference value" that is supported by actual measurements, whereas a without the OAL gauge and the measurements, it would really be nothing more than a guess. In theory, you could use a guess solely as a "reference value", but I think most would prefer to use the OAL gauge [measured] reference value that is likely very close to the true value rather than a guess. There are, of course, there are several other reasons why the reference value should be as close as possible to the true value (i.e. more accurate than a guess). Nonetheless, it's worth noting that there are certainly potential caveats to the OAL gauge reference value...to name a few:
1) the distance from the base of the OAL gauge case to the bullet ogive is NOT the same as measuring
from the bolt face
2) the OAL gauge case may differ in external dimensions slightly from a piece of brass fired in the rifle
3) the point on the bullet ogive where the caliper insert seats may actually NOT be the same point on the
ogive that first contacts the lands
4) some find identifying the very slightest touch of the bullet to the lands to be tricky
In using the OAL gauge, we assume these potential sources of error are relatively minor, and set the values generated with the OAL gauge as a reference. The good news is that once you have generated the distance to touching value with the tool (i.e. CBTO reference value), from that point on, you are directly measuring loaded rounds. Those directly-measured rounds are then fired, and optimized seating depth values determined from the target. It is then possible to reproduce the CBTO values determined from the directly-measured rounds with very good accuracy and precision. In other words, the OAL gauge reference value has effectively been rendered moot at this point because you are now measuring actual loaded rounds, and determining precision on a target using the actual measured rounds. You can then reproduce that exact CBTO value at will.
In using the CBTO determined at "touching" with the Hornady OAL gauge as an effective zero point (i.e. bullets seated longer are jammed and bullets seated shorter are jumped), it doesn't really matter if the true measurement to "touching" is different by a few thousandths. It's just a reference. Hypothetically, let's say someone pushed the bullet in a tick too hard, so that it really ended up .003" into the lands when they thought it was just "touching". Thus, an "optimal" seating depth of .012" off the lands as determined by the OAL gauge reference would in reality be only about .009" off the lands. But it doesn't matter. The resultant CBTO values from rounds prepared for a seating depth test would be measured directly from the actual loaded rounds, which then would fired, so that the optimum seating depth would be absolutely clear from the target, and easily reproduced at the reloading bench. In all likelihood, the reloader would go on thinking their bullet ogive was .012" from the lands, when in reality it was only .009" from the lands, and never know the difference.
