The Hornady OAL gauge/Stoney Point tool is designed to provide a cartridge base-to-ogive (CBTO) reference point, nothing more. From the established CBTO reference point at "touching", one can set up a seating depth test, using CBTO measurements from actual loaded rounds, which are very accurate. The seating depth test will then provide direct evidence on a target as to what is the optimal seating depth. Loaded rounds can then be re-produced as needed at that exact seating depth using CBTO measurements. In other words, once the reference point has been established, there is no need to rely further on measurements taken with the OAL gauge.
I typically select 10 bullets at random from a new Lot#. I number the bullet base with a Sharpie, and measure/record bullet OAL and whatever other bullet-specific measurements I feel like taking. The measurement set is labeled and stored for use as long as I am shooting that specific Lot# of bullets. When using the Hornady OAL gauge to establish the "touching" reference point, I measure CBTO and COAL with all 10 bullets, then take the average as my "touching" CBTO and COAL. Of these measurement averages, CBTO is the more useful in terms of reloading; i.e. setting up a seating depth test. I primarily use the COAL values and average more as internal controls to be certain I am not introducing unnecessary error while taking the measurements. For example, when using the OAL gauge correctly, I do not want to see a much greater SD for the COAL measurements than was obtained from the individual bullet OAL measurements. If that happens, it usually means excessive error was introduced while taking the measurements with the OAL gauge because variance in COAL should be directly related to bullet OAL.
I also use the initial average COAL measurement to set up a QuickLoad file. However, I sort bullets by OAL for the purpose of uniform pointing, so the average COAL measurement taken with the 10 bullet measurement set and adjusted for seating depth may not be the true COAL measurement with a specific length group of bullets. Further, bullets tend to "wobble" a bit when taking COAL measurements with the Hornady OAL gauge because the neck is pretty loose. It does not seem to affect CBTO measurements nearly so much. These discrepancies are easily corrected by determining average COAL from a few rounds loaded at a specific seating depth with bullets from a single length group, which can then be used to modify the QL file. Because they have been length-sorted, the bullets within a single length group do not typically exhibit the same OAL variance as the OAL gauge bullet measurement set, which were selected at random to better represent the Lot# of bullets as a whole. Further, pointing changes bullet OAL, so it is better to determine COAL once an optimized load has been established. Attached below as an example is a measurement I took with the Hornady OAL gauge and a new Lot# of bullets a couple days ago.
Measurements obtained using a Hornady OAL gauge can be very useful to establish a CBTO reference point to facilitate the load development process. However, I wouldn't get too far down into the weeds worrying about subtle differences in measurements between the Hornady OAL gauge case and fired cases from the rifle. Because one is simply trying to establish a reference point, these subtle differences don't really matter, because you can take much better measurements from actual loaded rounds that correspond to real-world results on a target. If it really bothers you, you can always make your own case by tapping a fired round to fit into the OAL gauge as was mentioned above. As I recall, the thread pitch necessary to fit the tool is a little oddball, but tools to do it are commercially available.
