The one curious thing I noticed when throwing the Elites on was the AF1's have more threads than the Thunders do and the threads appear to be a slightly smaller outer diameter than the axle itself. Even if you ditch the tube spacer and speed washers you're not getting the center race of the bearing to have enough contact on the axle itself to prevent it from having a free wobble, which I assume is contributing to the issue (impact on the wheel from it hitting pushes unevenly on the center race and causes the outer bearing to go crooked). This is something I didn't even have register to me as potentially being part of the issue, so we'll see once I get the Stage 4's if it makes any bit of difference.
Thanks for the detailed info - it sure is one of those funny ones, to happen to so many setup options, but as you came to the conclusion in the last paragraph, I am wondering if Ace "quality control" had something to do with it, not wholely the issue, but at least contributing to it in some way.
That was the thought I had earlier, but I didn't want to say it straight away, as well as not having your setup to look at or skate, so I didn't know all the facts and was limited by my own experiences and what I read for this one.
I used to push one back wheel out on a set in question on smith grinds on ramps, due to the angle and the pressure, so if I didn't do that trick, I would have fewer issues, but every run I would have to pop the wheel back on to the bearing correctly, if a smith was done in the run. Not being about to stop doing smith grinds, so I tried the same as race reds (different brand) and it stopped the issue, but my board did have a little of the "dead" sound from having that one axle nut tightened down, but I left all the others with some play.
When I switched out those wheels, I didn't have the same issue, but that was my experience there. I since put those same bearings in cruiser wheels that are softer with no core, and they have stopped the same thing happening when I use that board, hard sharp curvy turns would usually result in the bearings moving in the wheels, which now hold fast, again with all four tightened right down, but I am not worried if there is a little lateral movement as everything stays where it should be and the bearings spin freely.
I know that doesn't help this situation, but that is in detail pretty much my own experience with similar things. Others I have known to have wheels with bearing seats widening out, so the bearings pretty much fall out of the wheels usually result in those guys just stop skating those wheels, or have tried various bearings until something fits, even to the point of putting something in the wheel to try to gap up the difference, but I think that is not a good solution.
Wheels being urethane will bend and stretch even more, or reshape / misshape easily.
Also to mention here, there had been issues with some Spitfire wheels where their core was not made correctly, so one bearing might not seat fully, but the wheels I saw that had this issue were few and far between and I could not get bearings into the one wheel in question to full get it on the truck normally, so this is almost completely ruled out, even though I could see mention of quality control issues for some things like this being brought up.
The list of the big wheel brands all having the issue made me think of other factors, trucks maybe, or whatever, but any which way it sure is annoying to have that happen.
Not to ask too much, but is it possible to set up or have filmed a feeble or trick in question that you think is causing the issue? I don't care how good, how big, how well done it is - just want to see it, more because I am so curious, but also this sort of thing is very interesting to me in terms of knowing what and all that.