I've had rifles that required "chasing" the lands. I've had others that didn't and continued to shoot with the bullet seated at essentially the same original optimized seating depth as when the load was first tuned in in a new [broken-in] barrel.
For every Lot# of bullets I use, I remove 10 at the start, number them, and take all the pertinent bullet measurements. This represents the "measurement set" for that particular Lot# of bullets and I use them to estimate the distance to "touching" as long as I'm shooting that Lot# of bullets. By taking regular measurements every 200 rounds or so, land erosion is relatively easy to document over time. When the lands have eroded sufficiently that the original seating depth tune might be in danger of going out of the window, I simply do another seating depth test. If the results indicate the optimal window has changed, I seat the bullets accordingly. If not, I leave them where they were. I haven't been able to get any good handle yet on why a particular rifle behaves in a certain way with respect to seating depth, which is why I determine optimal seating depth empirically at regular intervals. The nice thing about seating depth testing is that we don't have to understand every single variable that may be changing when we optimize it. We do the test and hopefully obtain a solid answer where the bullet wants to be seated. As a friend has told me repeatedly, "we also want to know why things happen a certain way...but sometimes we just have to settle for what works".
I would make a rough estimate that maybe 2 out of three of my F-TR competition rifles require chasing the lands, the other third does not. My guess would be that the bullet itself has a lot to do with it (i.e. weight, ogive profile, bearing surface length), as well as freebore length/diameter, barrel land/groove configuration, etc. For example, I can imagine that a tangent ogive bullet might appear to want to stay at the same seating depth for a long time. However, that could simply be because many tangent ogive bullets are far more forgiving with respect to seating depth optima than secant ogive bullets. It's hard to know for sure. You mentioned the change in effective case volume, but it's also worth noting that when chasing the lands, the relationship between the bullet ogive and the rifling is not the other only variable that is changing. Over time, the bullet will also have less and less bearing surface seated in the neck, which could also play some role in precision. The real question there would be whether moving the bullet a few thousands at a time would really change the grip sufficiently to cause a noticeable difference in precision. The same can be said of changing effective case volume. In fact, the effect of seating depth on effective case volume can be effectively assessed using velocity as a readout. But again, the characteristics of a given bullet, chamber setup, whether the bullet is being jammed or jumped, as well as a few other factors are what likely make this topic too complex to ever describe in any manner sufficient to draw some general conclusions that extend universally across a wide range of different rifles.
Edited to add: as an interesting aside, I recently had a couple of my .30 Win F-TR rifles re-barreled with .040" longer freebore than before, to allow a little more case capacity with Berger's 208, 215, and 220 gr .30 cal Hybrids. Previous barrels were all chambered with 0.180" freebore, which is just about perfect for the Berger 200.20X bullet favored by many F-TR shooters. The new barrels have 0.220" freebore, which still allows the use of the 200.20X bullet, although they are seated fairly far out the neck. However, it allows for a little more case capacity with the longer, heavier bullets. In the past with multiple 30" MTU contour barrels having the 0.180" freebore, tuning in previous Lot#s of 200.20X bullets was almost like clockwork. Seating depth would typically be optimal at .009" and .012" off the lands. This has happened with multiple barrels and Lot#s over a period of several years. The groups with bullets seated .015" and farther would start to open up again, whereas .003" and .006" off were not great, either. In doing some load development with one of the new longer freebore barreled rifles, I was surprised to find with multiple seating depth tests that the optimal window with the current Lot# of 200.20Xs starts at .006" off. The interesting thing is that this is a new Lot# of bullets I have only used in one other rifle having the much shorter freebore length, where they also started to come in at exactly .006" off lands. So two different 30" barrels, with .040" difference in freebore length, both tune in with a particular Lot# of 200.20Xs at .006" off the lands. I appreciate the sample size of this observation is too small to lend statistical significance to this observation. Nonetheless, I've never seen any other Lot# of 200.20Xs tune in their. It could be some small measurement error in determining the distance to "touching", but I don't think so because of the measurement procedure I use. My initial reaction to this observation is that perhaps the ogive profile of this particular Lot# of 200.20Xs is somehow a little different than previous Lot#s, and they therefore like to be seated a little closer to the lands. Realistically, I will never be able to "prove" this theory, but having the new Lot# of bullets tune in at exactly .006" off the lands even though the freebore lengths are quite different suggests something intrinsic to the bullet itself may be cause, rather than the chambers, which have a significant difference.