Thanks for this explanation, it makes sense to me now. I was having trouble with a couple of the bullets sticking in the hornady modified case, which made it difficult to know if I was "kissing" the lands, or had to push harder to get the bullet to slide in the case, so that was part of my variability. The long ogive profile of the ttsx probably made it stick sooner in the comparator base, hence the longer reading. I wasn't thinking about how the difference between the diameter of the comparator and my actual lands would affect the measurement.
One related question though. Barnes says to load the 150 gr TSX to a COAL of 3.27. Based on the measurements I did, that would be a 90 thousandths jump, which is way longer than I would have started. But now I feel like I should ignore the CBTO measurements for the reasons mentioned above and just go with the Barnes COAL?