Problems with these notions,,
If not turning your necks with a high quality neck turner you can expect to see bad runout as well.
The root cause of case runout is thickness variance. It's brought out by sizing(to cause yielding). With this, if your cases have no thickness variance at any datum & 360deg around, then there is no runout brought into play. So necks do not have to be turned to be straight, and turned necks will not guarantee low runout.
If only sizing half of the neck, the bottom half could be out of alignment from the factory.
In a broad sense, we should never, ever, size the entire length of necks. There is only bad in it for most. The neck shoulder/donut area of necks will straighten out as well as they will ever on fire-forming. Otherwise, you bring out the demons lurking there.
Factory chambers can often deform brass.
I wouldn't say often,, have never seen it. Usually, factory or custom, your chamber is your best die.
I'd also recommend removing the expander ball because it defeats the purpose of a neck bushing die by only allowing one neck tension which is only the diameter of the expander.
Tension is not adjusted with bushing size(other than achieving nominal), but with sizing length against seated bearing. Nominal, as recommended by pretty much any bushing die provider, is 2thou under cal(as intended, with any bushing size providing it). This is actually nearly 1thou excess(no big deal) and follow-up expansion would ideally leave necks 1thou under cal. This is with spring back in both directions, and the spring back from ideal 1thou interference fit is all the bullet grip you need, and normally all you will get(regardless of further interference fit). The length of this spring back against bullet bearing is what pressure will overcome to cause bullet release. And also, using your bullets as neck expanders is a bad idea.
You should imagine in this that the greater the tension to overcome, the greater the variance in it. This is not good for all that's going on early in the powder burn toward your tune timing. This is why you should never FL size necks. That is, unless running a peak pressure node up around 75Kpsi in a tiny under-bore. At this point tension matters none, other than to hold the bullet(contain the charge) long enough(not the right amount of time, but at least an amount over .00xxsecs).
So it is possible to increase tension beyond spring back against seated bearing. FL sizing of necks
and not expanding them, will do this. Not shooting a 6PPC or 30BR? Don't FL size necks.
So why is neck pre-expansion good? It drives thickness variance outward -away from seating bullet bearing -reducing loaded runout (from one cause). It also leaves brass imbalance biased inward(countering last sizing action). So with time here, neck tension would rise rather than fall. Without expansion last, tension would decrease over time. Something to consider if storage is needed, because you may want it one way or the other. But also this creep is under ~1/2thou, not huge. If recently process annealed, that 1/2thou might mean less. All this is about achieving consistent neck tension, and seating bullets straight.
Button pulling through necks with success takes a plan. I believe forster has worked on the issues.
Another/better plan is to use Sinclair or PMA neck expander mandrels for this specific operation. Their mandrels are already sized right.
You can of course bushing neck size down, without follow-up pre-expansion. If your necks are turned, and your bushings are just right, you can get away with this without detriment.
There are qualifiers, or prices, for everything.