[Tyhm]

1) regarding the global quest: Yeah, I keep coming up against the same problem: a single quest on many servers, any one server can break down or rebel and screw it up. Thus, each server can only write to its own globalstrings and can choose to ignore the strings from others - if world A sets global.A.0=gotthegoldsword and world B doesn't care, they don't even have to script it. I look at it like playing a sequel on the Playstation: it reads your memory card, if it finds a save for the previous game you start with some easter egg unlocked.
It'd be cool to have a quest that requires that much cooperation though. Like a quest in Chrono Trigger, you gotta hop in the Epoch, dive into a dungeon to set things in motion to hop back in the Epoch...
2) I like the idea of paying developers somewhat, but as a former paid developer I gotta say it's cursed. There are expectations that come with a paycheck, and there's a finite level of enthusiasm you can put towards a company that's your source of income...we all dream of living large off a game we made, but it's the spirit of the hobbyist that made Graal great. Thus say I: make the tools easier to master, so the new and enthusiastic hobbyist is not dissuaded from making his custom system. I say start each world off as though restoring from a backup of an elegant yet primitive world-they have a serverside movement system, their own website where they can register accounts, a really basic customizable client frontend. It should clearly be Bob's Game, not Bob's Slice of Graal Estate.

Once it becomes easier to sell Bob's Game, it won't matter how much Graal's paying him to develop content for them: it's Bob's Game (hosted by Cyberjoueurs, powered by Graal4). At that point, Bob's working for himself, reaping reward and folly alike.