Nothing totally conclusive, but proving a negative is more involved than proving a positive. I’ll put it this way: I haven’t had any stressed stems I could directly attribute to any of my Gazzew switches.
I have traced a few to plastic-wrap shims, pre-retool Kailh Box switches, and even prototype umwipe stems - with a few being inconclusive as to the culprit. If I ever do see evidence of Gazzew switches straining stems I’ll document it here, but I haven’t yet.
Last I heard from Gazzew was similar; no evidence to indicate a problem and he’s comfortable calling them safe. The GMK set I’m typing on now with TTCs underneath was previously mounted on some U4s, and they haven’t had any problems that I can see.
From what I understand the spec is a hair larger than Cherry at its largest point, but not to the degree of even current Box switches. It’s also true that Gazzew’s stems have a taper on the cross-mount and are widest at the top, getting slightly more narrow towards the bottom. In a case where the fit is very tight, I think this shape lets the stem deform a little if it needs to before cracking a stem - just a guess there.