You said you have gauges now, what are you measuring the CBTO with? Asking because if you're using the Hornady/ Stony Point or Sinclairs you aren't actually measuring the surface that engages the lands. You're measuring an arbitrary point forward of the engagement point defined by the size of your comparator insert. As an example, my 30cal insert is undersized and not concentric with a minimum of .280" and maximum of .293" (as I suspect all of them are because they're so cheap). The result is you are actually not measuring a change in the beginning of the bearing surface where the lands are engaged, but rather a change in a point along the profile of the ogive. This is a case where your older method of actually putting a bullet in the bore is likely more accurate because it measures the actual interface point.
The comparators works because there is a fixed relationship between your particular comparator insert and the beginning of the bearing surface for high-quality bullets with tight one-lot manufacturing tolerances, but that fixed distance does not necessarily remain constant lot to lot. Even if there was some high end comparator insert bored to the exact bore measurement of the caliber to measure a consistent engagement point, that measurement would vary between chambers and over the life of a rifle. So make it easy by removing as many variables as possible and measure the actual engagement of the bullet to the bore in this rifle.
I would use your original method to correlate the comparator CBTO measurements with actual lands engagement, and use that, not 0.009", as your change. Since you aren't at a max load, I'd probably just shoot a bracket around that new measurement, +/-0.005", and call it good.
I thought the whole point of measuring the ogive was to measure the relationship between the bullets first point of contact and the lands. The CBTO measurement doesn't account for length of the bullet, bearing surface, or boat tail, all of which will make bullets behave differently. CBTO is...
www.longrangehunting.com