Of course, having to test all of the units (for calibration, at least), they will mark as 50MHz the worst ones, but there could be potentially good 100MHz ones sold as 50MHz.
Absolutely; I'm sure that there are.... and perhaps even more now since I'm sure they've had a surge in sales since this hack became public.
But one thing I've learned from most Chinese companies is that they might do anything to save a buck if it doesn't immediately and directly impact sales - such as overdriving circuits, not providing any serious support, not fixing bugs, etc.
So if, pre-hack, Rigol was selling, for example, X number of DS1052s and X/2 number of DS1102s - their design and choice of components would reflect, in general, the need for the amount of tested scopes to fall along that curve.
Now let's say, post-hack, they are selling X*4 number of DS1052s and X/4 number of DS1102s, and they could save X number of dollars by substituting slightly less costly caps or speed-rated versions of ICs because the tested scopes will still fall along their new sales projection graph; would they do so? Perhaps, perhaps not - it's hard to say.
In any case, I will likely do the hack after my warranty runs out
(as well as changing the fan and adding a battery pack). I just will be aware that it could be slightly lacking in some respects to a tested-at-the-factory 100MHz scope.