Ah, it's fun to watch as the half-truths fly on "neck tension" and collet vs. bushing neck sizing is discussed.
First and biggest thing that's missed is the inside diameter of a neck is only a moderate part of bullet grip. What's called "tension" is what machinists call an "interference fit". Any neck diameter smaller than about one thou is meaningless because the brass simply deforms as a bullet is seated. Within the elasticity of that one thou of interference fit, the thickness and hardness of the brass itself controls bullet grip and it varies by individual cases within the same box; we can easily feel it in the variation of seating resistance from round to round of carefully neck sized new cases. That uncontrollable difference in the cases easily negates all the agonised anal care in finding just the right bushing for just the right "tension" of each neck.
The spring back from normal brass work hardening will slightly change the final inside diameter of collet sized necks each time they're cycled. Exactly the same thing will happen with bushing sizers. About all we can do about it is to anneal after each few cycles. And it's likely more cases are destroyed for fine accuracy by annealing than are helped - annealing just isn't all that easy to do precisely.
All a bushing die can do is insure the outside diameter is close to the bushing diameter - that's generally about a thou larger than the bushing. And all that means is any variation in neck thickness WILL be pushed inside and the "neck tension" will vary that much, round to round, UNLESS the necks are turned full circumference and to identical thickness. For BR people who have chamber necks that demand turning necks thinner just to chamber the cartridges, that's fine and it was for them bushing dies were first made. SAAMI chamber necks are oversized enough that very much case neck turning makes an already sloppy chamber fit even sloppier.
Lee's collets provide straighter necks with more consistancy than many of us have ever achieved otherwis, and does it with any cases. But it's at its best with lightly skim turned necks to eliminate the worst of neck thickness variations. And the inside ID is always quite consistant if not perfect. Anyone thinking Lee's collet will squeeze one side of a neck down more than the other hasn't a clue.
Leaning harder on a Lee collet die can/may reduce the necks maybe another half thousant. That's enough to make a different feel of seating but I haven't seen it affect accuracy. And pushing harder does present the possibility of stripping the aluminum top cap right out of the die; that's not a flaw, it's a carefully designed protection feature to prevent serious damage to the die body or press.
Bushing sizers were invented by BR guys needing to deal with tightly fitted chambers in their single shot rigs requiring highly processed cases and very low bullet grip. But, as so often happens, others happily figgered if a bushing is good for BR it must be great for them; as with a lot of such blind 'follow the accuracy leaders' ideas, it's really not true. Their rigs and ours are much different and our needs aren't well served by following their very specalized paths. But the DIE MAKERS are quite happy to provide anything the market will pay for so....buy what you want, it's out there just for you!
Lee's collet dies were made for the rest of us and our SAAMI chambered rifles.