AFAIK Twitter already considers bots impersonating people (if not just all bots) to be against their TOS and actively works to hunt them down or remove them. The fact that they're imperfect and so your strategy might work some of the time has nothing to do with the proposition that OP is making.