FWIW I only ever neck size. I haven't full length sized any brass for longer than I remember. In fact when I first started out, nearly 40 years ago, neck sizing was all I was equipped to do.
I reload for a variety of firearms, including break-action, lever action and falling block, as well as bolt actions with front locking lugs and rear. To the extent relevant some use rimmed and some rimless.
I use a mix of Lee Collet dies (which I prefer) and others.
I seem to get good case life - have batches of brass that went past 30-odd reloading cycles before I stopped counting, without ever being FL sized. They do need necks annealed every now and again though.
It seems to me that the greater degree of cold-working inherent to FL sizing may not augur well for case life (though of course you can anneal the neck and shoulder where the bulk of the working occurs), but other than for .303 in my Lee Enfields I haven't run the experiment. With the .303s I found that FL sizing had a definite adverse effect on case life, but that is a bit of a special case: these chambers are cut pretty loose, and generous in the headspace, and so if you FL size you tend to see not only a lot of deformation but also stretching and, quite quickly, case separations. I avoid problems in this regard by taking care to fireform symmetrically, by using a narrow strip of tape or o-ring around the case in front of the rim, so that the case will be pushed hard back against the boltface and expand to fill the chamber. After that I neck size only, and the case headspaces off the shoulder not the rim, avoiding the stretching problem.
I have often read people talk about the need to "push the shoulder back a bit" after a few reloading cycles, or to FL size to ensure that the case will chamber. I've never really seen a need for either. It seems to me that if a case came out of the chamber it simply has to fit back in again, unless you do something to it in the meantime. In fact, given that brass obeys Hooke's law within its elastic limits, the case is always going to end up smaller than the chamber, and it sure can't end up bigger.
The only two likely ways I can see why a case you fired in a chamber won't go back in again easily are either you did something to it which altered its dimensions, in the reloading process, or the chamber was loose or out of round, so the case is no longer symmetrical, and won't fit back in unless oriented the same way as it was when fired. I've only seen evidence for the second one in very loose chambers. I suspect that more often than not it is a result of the sizing method, and that the need to push the shoulder back every now and again is because you were inadvertently pushing it forward due to your reloading process, such as by the effect of partial sizing or by dragging from an expander.
