You're talking about a coup or revolution. If you want to say that, I think the efficiency of a coup or revolution doesn't vary according to the government system that is getting attacked. The efficiency at which such attempts operate are a function of how organized they are.

No, I'm talking about an unpopular dictator simply resigning (because he is unpopular) and appointing his replacement! Highly efficient.

I guess hence your qualifier "whether it's likely". I'd find that even less likely than a revolution, but I have no idea what the real actual statistics are. Would be interesting to see data on coups vs resignations vs stayed in power until death or retirement.