If he's heated it up too high it might be even harder than factory, depending on the hardening response. Reynolds said that you shouldn't drill after welding or brazing because it would be too hard, so high heat may result in even harder (and less ductile) than factory. It seems to be somewhat supported from the fusion welding in the Reynolds chart showing an increase in strength.

There's also a limit to how much steel can be cold worked before failure. The stays were already squished some, whatever you do is likely adding on top of that. Normally steel is annealed to reset the microstructure into a soft state ready for more cold working, like during the butting process. Heating an air hardening grade that high will result in hardening instead, and even if the stresses and microstructure are reset there, the whole tube isn't being heated. I don't really know the specific behaviors of 853 under heat treatment, but obviously, real actual controlled heat treating results in better results than waving a torch until it seems nice and hot for various reasons.

Sorry the advice isn't more practical, but this is clearly going against best practices, so everything is compromises and guesswork. I don't really know what the minimum you could get away with would be. Thinking on it some, first I would try to avoid this situation altogether. Since it's too late for that, I would probably try to temper it around ~750C, around the point when steel will stop being magnetic. This is pretty much critical temp for some steels, but Reynolds says temps lower than 850C will soften it, so I'd hope Reynold's metallurgists are relaying information accurately. I'd crimp until happy, with as gentle a form as possible, then bring it up to just north of 850C to minimize grain growth, maybe up to the point where bronze filler just starts melting but not further. Hopefully this deals with any of the stresses at the crimp, and makes it nice and hard to help resist abrasion from tire rub or nicking from chain suck.