Another test I would love to do (as much as it would likely not even be rigorous and would only be a single data point) is to run back to back two sets of PS2, one say <10k mi, <6 months old, around 50% worn, vs. a new then immediately shaved set of PS2's, shaved just into the wear bar or say slightly above the wear bar. I think their is ample evidence and opinion here that they would be very close in traction or perhaps the latter having some advantage. I think the only way this could be the case is if those of us who have experienced significantly observable traction loss under a similar extreme wear scenario are also feeling significant negative/confounding heat/UV/aging effects.
As with many complex things I examine I simply like to see consistency among anecdote, testing, theory, common sense, etc. and this one is vexing...