Gamefreak have tried before to design insanely powerful pokemon with a shockingly bad ability (Regigigas, Slaking, Archeops) and they’ve generally failed or at least turned out pretty shakily. Do you think it would be possible to ever do this and make a consistently usable pokemon?

Hrrm.  Tricky.

I feel like it has to be possible, because we have items that give their users severe disadvantages, and those get used all the time.  If you imagine for a moment a Pokémon with really high attack and special attack scores, maxing out somewhere in the 550-600 region, and decent stats elsewhere, whose ability locks it into using only one attack until it switches… people would almost certainly use that, right?  Because people use Choice Band and Choice Specs, and that’s basically what I’ve just described, in ability form.  Obviously this particular example is impossible because if a Pokémon like that existed, people would stack choice items on top of its existing advantages for no extra cost, and all hell would break loose, but the point is that some disadvantages are clearly worth it.  The trouble is that I’m pretty sure the competitive multiplayer environment is not really on Game Freak’s minds when they playtest these things, and that’s the only way you’re ever going to draw the line between broken-because-good and broken-because-bad, both of which will be serious possibilities whenever you create a Pokémon like this.  I mean, Archeops is fine in single-player; I used him in my first playthrough of Black and he was fantastic.  Slaking has the potential to be ridiculous in the right hands because the AI doesn’t know how to exploit his weakness.  So the designers are kind of firing shots in the dark here, I think.  That makes it unlikely that they’re ever going to get it exactly right, but sooner or later they’re bound to get something that falls on the overpowered rather than the underpowered side if they keep trying.