I have, myself, experimented with concepts involved with moving some of the unknown into the hands of the players. While I haven't reached a satisfactory solution yet, as far back as this, I have toyed with this idea.
When viewed by the recently popular 'allocating credibility' perspective, it follows that ultimately the typical game revolves around a conflict between player and GM, who gets 'the say' about what happens. This is because GM has what players want and they receive it only at his discretion. Arguably this means that only the GM's word had the final 'credibility'.
If you look at typical gaming in more of a 'spotlight manipulation perspective', the 'power' shifts back to the players' hands. Employing features offered in chara generation and play, players are able to take the spotlight when and where they wish, even though the unknown remains the purview of the GM. I've heard of this being called things like 'getting to be the bad-ass'. Either way, in typical gaming, interest in the game generally rises out of novelty and curiosity about the milieu.
My interest lie in exploring the effects of a 'spotlight' based perspective rather than the 'credibility' model, largely because it doesn't require any more 'sharing' of GM responsibilities than typical gaming did¹. The whole process of sharing the GMing destroys very much of the novelty. In the past, I was frequently guilty of forcing the unknown on unsuspecting GMs through things like amnesia or unexpected enemies. So I'm just like that.
¹ Maybe I should talk about 'stealth GMing' used players in typical games to this day.