This is exactly the type of anecdotal evidence that pervades oil (and a lot
of other) discussion, as I'm sure you know. Causality in general, in any
subject,  is EXTREMELY hard to prove. Just because you can show correlation
(between, say, oil brand and engine longevity) does NOT mean that the one
caused the other.  One would think it does, but it does not. Why? Because,
among other things, there may be a third factor (say, those who use Brand A,
ALSO keep the rpm under 4,000--or whatever) that accounts for the "effect."
And with oil, we are not even at the point of showing true correlation. This
logic is something that medical researchers take for granted. The standards
are far more relaxed in the automotive world.

I personally am a lover of all those snake oil elixirs on the FLAPS shelves,
not for the logic but the voodoo of it. Do they work? Most of them, hell no.
But they allow you to dream a little bit, and that's what it's all about,
right? Otherwise, what would we be doing driving these old hippy vans anyway?