If you use a Stoney gauge, or something similar, you will always know what your seating depth is, regardless of the COAL.
Well ... no, while you are right in theory, I disagree in practice because there is often significant variation in ogive shape from bullet to bullet within a box of bullets. This can result is as much as 5 or 10 mills of variation in distance off the lands for some bullets.
Any gage system that measures to the lands only tells you what the base to ogive distance is for that particular bullet used to measure it. Take a box of bullets, measure the base to ogive distance variation on the bullets themselves and you may be surprised to see how much variation there is in some bullets. That translates into variation in ogive to lands with the seated bullet unless the bullet seater in the die is shaped exactly like the lands in the chamber (made using the exact same chamber reamer as was used to cut the chamber).
I'm not convinced seating depth variation within those limits is important, but for those searching for the "exact" same distance off the lands for all cartridges loaded with bullets from the same box, it will be frustrating because it is almost impossible to achieve in any practical sense.
The COAL might be a ballpark guide, but since bullets differ, even within the same batch, it seems to me that it is not an accurate way to know their relationship to the lands when determining seating depth. Am I wrong on this?
COAL is probably close enough for bullets that don't have to be jammed. One can measure to the ogive for quality control, but adjusting the seating to make them all the same is going to be really difficult with out using some sort of custom die.
For the reasons you have stated (variation in bullet ogives and tip distances), the only way to know exactly where your bullet is relative to the lands is to "lightly" jam it in. For some bullets, that results in more accuracy than seating it to be someplace around 0.020" or <insert your favorite distance>. For other bullets, it doesn't.
I check several bullets and use the average with my Sinclair tool to establish a COAL that will give an ogive to lands gap of ~0.020" because I've never measured a box of bullets that had enough variation to allow them to contact the lands and cause a pressure spike when seated like that. It's the least jump I'm comfortable with other than when the load is developed specifically for lightly jammed.
A bullet seater that pushes on the ring of contact where the lands would touch the bullet might (probably would) deform the bullet and probably screw up accuracy (and distance to lands due to distortion of the ogive at the point of contact) more than seating depth variation that occurrs due to bullet variation. So one lightly jams it and hopes that doesn't distort the ogive enough to cause accuracy problems.
The only bullets I shoot barely touching to lightly jammed are 168 hunting VLD's in my 7mmMAG (they are the only VLD's I shoot so far). The VLD bullets have such a long nose I think it's important to have them exactly centered in the bore and lightly jamming them does this. But the load was developed from the beginning with this in mind. Jamming may cause pressure to spike several thousand psi higher than it would if the bullets had at least some jump.
Every other bullet in every other rifle (except my CZ Hornet which is magazine limited), and that rifle, is seated about 0.020" off the lands because that's far enough that the +/- the variation in factory produced bullet ogives won't allow them to touch.
Trying for exactness (variations of less than .003" to .005", or "just" touching) with production bullets and brass is like using a micrometer to take quality control measurements for precision cheese slicing.
Finally, worrying all this to death, and getting the base to ogive distance exactly the same on all bullets, assuming that was practiacly possible, with out measuring and fixing variations in eccentricity is a waste of time. I have a heck of a time getting the ammo for my 7mmMAG to be straight (have zero run out on an RCBS Case Master). Using an RCBS competition seating die I still end up with ammo that measures from ~0.000" of runout to 0.012" of runout. The CaseMaster is a wonderful tool for feeding endless frustration unless you also have a way to straighten cartridges. I purchased a small tool that looks like a piece of stainless steel angle iron with a bunch of what appear to be reamed holes in it. I stick the neck of the bullet in the best fitting hole with the case oriented so the bullet is pointing down and pull up on the base. With just a tiny bit of practice I've gotten pretty efficient at getting the runout to under 0.002" (+/- 0.001") first time.
I recently read an article where the author demonstrated an improvement in accuracy by straightening his cartridges in this manner which pretty much matches my experience.
Fitch
PS: FWIW, I bought the CaseMaster specifically to measure case wall thinning just above the base after having head seperation in my .30-06 ('53 Win Model 70). The load is nicely within manual maximums for the bullet and showed no pressure signs, but the case was stretching a lot resulting in an annular reigon of significant thinning after 3 reloadings just where the base taper joins the straight wall.
Resetting the sizing die to bump only 0.002" and lightly polishing the chamber of the rifle significantly extended brass life. Being able to measure the thinning allowed monitoring of the situation and enabled setting a service limit that eliminated case head seperations. But the only way found to measure it was with the CaseMaster. I don't know of any other commercially available tool for making that measurement.
frw