I've taken some time off from MMOs, going back to single player games has given me a great deal of perspective, in my opinion. Single player games have a few simple rules that seem curiously absent from MMO discussion. A: If this isn't your game, go play something that is. B: If you aren't in the mood to play a game... don't play it. It'll still be there tomorrow, and probably a lot more fun then. C: You're game's competition is not Starcraft, Brain Age, and Halo 3, it's whatever else in your genre is happening to release in a nearby time frame.
I certainly realize those don't work on a one to one basis in the MMO market, but they do help pad out some foundation. Before I get back into seriously discussing MMO design I'm going to lay out some basic rules, definitions and philosophies. I'm tired of talking at cross purposes with everyone, so from now on these can be used as a filter for all discussions unless explicitly stated otherwise.
The Definition of Success: Actually there are two. In the case that a core overriding goal is specified for the design, success is closeness to achieving or expressing that goal. If however there are no defined goals, a successful MMO has a subscriber base large enough to theoretically pay back it's development team in three years given it's business model. When in doubt, assume 30 million in development and a sub model of $15/month. That's about 220K players without box costs.
A Game Not the Game: The WoW killer and the Platonic game are pretty much pure flights of fancy at this point. If you find yourself wondering whether a design will kill WoW or be "The Perfect MMO", stop yourself right there. My goal will always be to design a game, one intended to exist in a field of many, not some ideal of a game. Also the mechanics will have flaws, period. You cannot design non-flawed mechanics, so any discussion of mechanics will be mainly about how well they fit the overall concept, not whether or not they are flawed.
When in Doubt, Quit: My personal philosophy is that if a game stops being fun, stop playing it, at least for a while. This also means that you can assume any game I design is supposed to, assuming I don't forget, encourage that behavior.
Two for Tuesday: A game exists within the context of the games around it. This also means that there is no point putting time and energy into a system that games around it simply do significantly better than my concept will allow. In other words, no "one game to rule all play styles" reasoning. If someone else does one bit better, I'd rather they played that for that bit, and my game for my bit.
The Inefficiency Coefficient of Efficient Content: Measuring the efficiency of content in number of players who get to see it is simply wrong in my opinion. Content does not exist on it's own and for it's own sake, only as part of a larger context of the game. Ergo if the game seeks to benefit players going off the beaten path, one time content is perfectly applicable to that. Content that is seen by all players but breaks suspension of disbelief or disrupts the flow of game play should still be weeded out. If you absolutely need some other measure of quality, then go to the level of positive reporting by the audience who do see it.
Ludic and Narrative Harmony: I strive for it.
Quest Hubs, Quest Givers, Quest Vendors, Quest Aerospace: I'm not going to say I won't use them, but it seems to me there is a lot of space to explore in what exactly most of these are.
Dude Where's All the Rats: Just because almost all of our games have rats doesn't mean you should just assume the ones I'm talking about do. Some won't have mobs, others just won't have static mobs. For even some others, they may be enemies who simply don't fit the mob classification.