Some years ago, I was far enough into wondering about this subject as to almost make a suggestion about it, but Arawn pretty well summed up the issues present. The root of the problem is that DMs have to have some distance from much of the PC action when they are on their PCs. The most obvious detail is that DMs aren't allowed to be in factions with their PCs. It's very hard to make a complete distinction between what you accidentally learn as a DM and what your PC knows - even if you can do that distinction, it is impossible for other players to be sure of your ability. Certainly, I trust the DMs and I trust the devs to be able to do distinctions like this, but if I didn't know them as actual persons, how could I ever really be sure?
Without full DM powers, some tasks are pretty hard to do, like restoring lost items. You need to have access to a bunch of logs and communication that can be of the sort that can spoil a lot of the in-character happenings. I used to be a DM for a while and one of the reasons for myself to leave the DM team was that I wanted to do immersive, sometimes conflict-prone stuff as a PC. It's hard, really hard, to do that as a DM, and still retain any element of genuine surprise in how events unfold.
The CC team can do some adminstrative tasks such as allow PrCs. I guess what could be technically possible would be to extend some adminstrative capabilities, such as restoring player passwords or booting players, as PC widgets that could be wielded by appointed PCs, but there are some problems there. Security, the like, and it'd need an application process of its own that would have significant overlap with the DM process.
I don't know if any such idea would be worth it, but personally I wouldn't be unwilling to discuss specific suggestions. I'm unaware if there's ever been a larger discussion about it amongst devs or DMs.
But in the end, even if there was (and I'm not saying there is) any kind of a consensus about such widgets being doable, it'd still hinge on someone actually going through the trouble of creating them in a way where we can be pretty sure about the lack of any significant potential for abuse. Unfortunately often, the question becomes about if it is worth the trouble?
