Coprolit alerted me to an article on Gamespot about the length of games. Basically indie game developers say that they can make games as good as anyone else, but due to budget restraints the indie games will be shorter than games from big companies. So Coprolit thinks that this principle could be applied to MMORPGs, with better gameplay instead of more content leading to longevity.
I'm not so sure that will work. Not that I'm against better gameplay. But being an explorer in the Bartle terminology, discovering new content is important to me too. Game mechanics *have to* be good anyway, because whether you kill the same mob over and over, or a different mob in every combat, the basic combat steps are repetitive. But players simply prefer a variety of different mobs to fight with. I remember in Final Fantasy XI when it came out there were very, very few different monster models. You left your starting city and killed rabbits, and when you travelled to higher level zones, many of them had rabbits of higher level, same model, different color. You could theoretically go from level 1 to the level cap by killing nothing but rabbits. That is cheap to produce, but not very interesting.
Coprolit quotes the problems of vertical, content heavy, level/area-based expansions, and I agree that I'd love to see more horizontal character development. That is character development that does not add to your effective power, but only to status or the completion of collections or better social contacts. But I doubt that could be done without adding content, the content would just have to be of a different kind. I've played enough games where the landscape was randomly created, and even some games where the quests came from a random mission generator. In general I found these games lacking in interest.
The one "indie" MMORPG I played was A Tale in the Desert, which is very good, but failed to keep me interested for long stretches of time. I think the huge mass of content that World of Warcraft has compared to its competitors is a big part of its success. That this costs many millions of dollars to create is unfortunate for indie developers. But I don't really see a way around it.