The issue regarding the amount of work required of the Delegate in the Delegate-as-leader system is a valid concern. The standard response is that officers and underlying bureaucracies (i.e. a group of people working under the Del that helps him handle the in-game regional stuff) will be apportioned much of the Delegate's duties, although responsibility for the carrying out of those duties will ultimately rest upon the Delegate. The success of a Delegate-centered system is very contingent on the quality of the officers that the Delegate keeps.
As for your second concern, Grosse, I am thoroughly confused. Do you mean constitutionally or in-game mechanistically? If you mean constitutionally, what keeps a delegate from going rogue? Nothing. (Well, influence does, but nothing can keep the Delegate from at least trying to go rogue.) The fact that off-site legitimacy has no standing on in-game delegate standing is one that we cannot avoid. No amount of red tape will keep a rogue delegate from attempting to go rogue. All that we can do is remove a rogue delegate's legitimacy through the powers of censure and impeachment. In my proposal, a Delegate's actions can pretty much be declared de facto illegitimate temporarily via the CLO and permanently via Assembly action. So, I have no clue where you're coming from here.
As for mechanistically protecting against a rogue delegate, well, that is true. The closest thing to that that I can think of is the maintenance of a body of high-influence citizens. It saved us in the encounter with Dalimbar and provides a high degree of security against rogue delegates (but of course also makes it easier for a high-influence citizen to carry out a coup). A little while back, Flem was talking about formalizing this body, although I would prefer to leave it out of the Constitution (and maybe even Legal Code) myself.