Thanks for the pointer to Powell's summary. I'd say #3 isn't very useful for our purposes since we don't use PO2 values over 1.6 bar, as well as Powell himself hedging his position because of vasoconstriction. #2 seems to be a misuse of the term to mean "inert gas gradient", so not terribly useful either. (No slight against Powell, as he's just reporting one way that term is (mis)used.)There is a summary table on page 135
So #1 seems to be the only real consideration for current diving practices. Lower tissue tension is further from the critical supersaturation level, yielding a reduced tendency to bubble. OTOH, we already have the widespread understanding that deeper has less tendency to bubble. Since the only way to widen the o2 window is to go deeper, it's kind of six of one, half-dozen of the other.
I suspect the term "oxygen window" has fallen out of favor because "deeper" is FAR easier to understand and yields the same benefit.
There's also the fact that the M-Values of the deco models were empirically set, which would have necessarily included some degree of the window-induced tension reduction.
Ultimately, we're still left with the trade-off that deeper is less likely to bubble but off-gassing at the associated elevated inspired PO2 may be impacted by pulmonary stress (vasoconstriction & mucus buildup). Therefore, whether 6m or 3m is "better" will vary.