... I still think there's an error there somewhere as the tighteneing should not nip the bearings. It's probably why the 2nd set shagged so quickly. ....
What it suggests is that there had to be some 'space' between the spacers and the bearings - the spacers should be a 'perfect' fit, so that when the axle is torqued, they do not impart any side-load on the bearing inner ring with respect to the outer ring.
Torquing it was then physically moving the bearing center races, taking up that space
Two causes of that
could be that the spacers themselves are not dimensioned properly OR (more likely to be honest) the bearing was not fully seated in the pocket.
That would then create a side-load once the axle is torqued in compression and cause rapid wear.
Did you happen to notice Steve, particularly on the long spacer, if it was snug? or loose? when you removed it?
And when re-installed?
I'm sure you've replaced bearings a number of times, so this is not meant to be teaching 'how to suck eggs':
best installation method is to 'press' them in - the press can be home-made, threaded rod with a couple of end plates;
I have a lathe so could turn up some plates pretty quickly. (but if you have access to some big sockets, those can work nicely)
the press should never be working on the bearing inner ring alone.
One thing that is curious is that the 6204-2RS is a sealed bearing, yet there is still an external seal used (item 4)
Was the original OEM bearing sealed?
Don't see it being a problem to be sealed but that makes the external seal somewhat questionable