In grad school I did a designed experiment and using statistics proved that.....in pistol rds anyway....bullet seating depth is the number 1 factor in accuracy.  I used different primers, powder, didn't matter....the key factor was seating depth.

And to add to your thought matrix....

Have you considered which data point to prioritize, ogive to shoulder datum or ogive to base of case?

In bottlenecks the former may be more important as the shoulder stops the case with the button ejector maintaining pressure.  (talking rifles and rifle cases with shoulders of course)

Pistol cartridges are more prone to pressure increases relative to seating depth so that may have more effect than with rifle cartridges.  I wouldn't give .004" a second thought unless perhaps I was shooting 1k bench rest.

Some guys think they see nodes based on 0.003" graduation (and I'm remembering/assuming <.001" variation). I think they're wasting components.

Would a prepped case with 0.219" ID make a decent 0.224" ogive comparator?

I shoot for less than .0005" between short and long with most being dead on as far as rifle brass, but I use RCBS powered lathe using their 3 way cutter which is known for accuracy, I also measure and set everything from the case-head and never the shoulder, the shoulder when used is just stacking inaccuracy.....I also use Berger bullets for most loads, rarely use Hornaday bullets, because they are so inconsistent...All of my dies are the good sleeved comp dies that help hold the bullet and brass square while seating bullet...I normally use Redding bushing die for neck only, and shoulder die for shoulder only, I also use a mandrel die for final neck sizing..I like keeping every action alone because it makes it easier to concentrate  on one at a time for more consistency.. That covers sizing brass, as for bullet seating I use forester sleeved seating dies for seating...they do an excellent job of loading consistent reliable ammo...