Isn’t most runout caused by brass thickness being uneven?
Many, many factors contribute to runout. With varying neck thickness, you cannot eliminate ALL runout. Additionally, you may find runout in different places so it is sometimes important to consider if there is runout where the cartridge fits the chamber, or runout where the bullet lies in relation to the cartridge itself. For instance, a cartridge fired from a given chamber should have next to zero runout in the neck taken from the outside circumference of the neck. Once it is sized, there will be runout due to neck thickness variation. Your technique may affect this greatly. Are you using an expander ball? How do you lubricate? Are you using a bushing or a fixed die?
For instance, using a floating bushing may induce slightly more than a fixed die because of tolerance stacking. In my experience it will amount to around 0.001" added after sizing if I am not using an expander ball. However, neck thickness variation is being pushed INWARD such that it will affect the bullet when seating. This may result in actually bullet runout of 0.003". What matters more, a bullet completely centered and straight to the case and chamber or a neck perfectly centering in the chamber? If I run that same piece of sized brass through a pin gauge or expander die, I may change that neck to runout to 0.003" but TIR taken at the ogive may be 0.002".
Which is more advantageous for precision? Well, it depends. Do you have a tight neck chamber? How much clearance does your neck have? Where are you in that case's life? Have you reached the point where the doughnut has expanded and it contacts the chamber's neck at the last portion of the neck that remains un-sized? Where is the shoulder of your case relative to the chamber? If it varies with every shot, how does that affect the seating node of the bullet?
In my experience (which my understanding of is evolving all the time), I've won a few PRS and F-Class matches, it is most advantageous to control as many variables at one time as you can. That means first, using good quality brass. Second, that means you want to do the same thing every single time. If you do those things, runout, within reason, has a minor effect on the rifle's performance. Typically, I don't even bother checking runout on a squared-away setup because even 0.005" TIR won't hurt performance when everything is done right, though its typically a little less than that.
My typical competition loading process is to use Lapua or Alpha brass and to anneal before any sizing takes place and do so every single time. The first firing or two, typically brass still expands and bumping the shoulder back probably isn't helpful but its totally possible that some random cases will expand to do so so the notion of not FL sizing cannot be ignored on a large batch of 100--there may be a few in there that will have difficulty chambering. Sometimes it isn't a bad idea to take a piece of brass and fire it two, even three times until you learn your chamber dimensions to determine how far shoulders should be set back before you do a batch of once-fired brass. Figure out that dimension and use that number with your comparator set. Shoot for 0.002-0.003" of set back. I do use a bushing without an expander ball, because I don't use custom sizing dies and I do not turn necks. For sizing lube, I will use Imperial and only lube the body and outside neck of the brass, I keep it off the shoulder. Following sizing, I use Neo Lube #2 inside the necks for lubrication and run through a separate mandrel die as to not affect concentricity or shoulder set-back as an expander ball being pulled back through the case may. Afterward, I trim, chamfur, and deburr all in one step with a motorized trimmer. These steps are quick, easy, and they produce the same dimension every single time. Consistency adds more for me than constantly varying brass and headspace dimension could ever do; especially considering the drawbacks of neck-only sizing as mentioned in previous posts.
