Hi Taff!
Thanks a bunch for taking those measurements. I'm baffled, because they are the reverse of what I expected; I expected if anything, mine, as used, to be worn and thus smaller.
At least they confirm that the walls were indeed different thicknesses.
FWIW, my dimensions are
thin = .48mm
thick = .89mm
overall = 3.46mm
But no matter, I am agreeing with your offset motion thing. Along the rear HALF of the perimeter of the cylinder/piston contact surface (call it from 12 o'clock to 6 o'clock) it is completely untouched. Whereas on the other HALF, (7 o'clock to 11 o'clock) there is some mild scoring. Not perceptible to my nail or a fine O-ring pick - but visible nonetheless. So, yes, for whatever reason, that piston has seen some minor offset movement relative to it's rotation center.
Also, and perhaps related to this, the thin edge seems a bit sharp - so I could see it causing / contributing to that scoring if the piston is tilted as you're suggesting. The O-ring itself looks happy and round (both in shape and in cross-section.
I'm going to mull your overall "stack" dimension thing. Maybe open up and measure the plates/steels. I'd like to know more about this "pill". Perhaps extending my overall stack is the reverse of what I need, but I feel like I'm slowly zeroing in the root cause, and it looks like there's a fundamental dimensional issue going on along that overall system.
If the "pill" is an item you sell, please pm me and we talk offline about the feasibility of ordering one.
cheers,
c