Note that the case neck brass hardness will also determine how much spring back it has from being sized down.
Won't this depend on whether or not it is annealed, (and how recently), as well as how many reloads the brass has had, and it's quality (e.g. Lapua or Win).
If one thou works now, might not two thou be the "right" number down the road? Just sounds like it's a variable, that's going to be continually changing.
So how to decide what's optimal?
No device I'm aware of to measure neck tension (pullout), is there?