I posted a comment on Tobold's blog for his Open Sunday extravaganza about smaller MMOs and decided to epand on that idea.
The current crop of MMO failures don't seem to be failing from their core ideas, but rather the implementation of too many features. They seem to suffer from a "grand scheme" syndrome. They came onto the market claiming to be able to please as World of Warcraft pleases, but would not be clones of the behemouth. Yet, inevitably they were just that.
The problem isn't so much that they looked like World of Warcraft, or even played like it. The problem is that they tried to put as much stuff on the table as World of Warcraft did. World of Warcraft is HUGE. After 2.5 years of playing World of Warcraft, I still have yet to explore every zone and am still encountering new aspects of the game for the first time. When Blizzard set out to make World of Warcraft, they obviously set out to also fill their online world with everything they could conceive of. Yet, even World of Warcraft was technically never finished even after all of this time. There are areas on the world map that are closed off, instance portals with no instance, battlegrounds with no battles, instances with no entry... World of Warcraft succeeded in spite of this because they had so much content that was complete and polished.
The recent crop of failures failed to take into account some important key details. The worlds they are creating need to be filled with appealing content that is polished and functional. If the world map is as large as vanilla WoW, but the content is rushed in to fill that world then you will end up with sparse, buggy features.
If the preliminary design meetings consist of grand ideas of gargantuan worlds packed with revolutionary features, it's already time to shut the game down. It doesn't matter if this game is as far from Blizzard's fantasy game as you can get, you are still trying to follow the World of Warcraft model of HUGE and PACKED WITH FEATURES.
Maybe Hellgate: London would have lasted if the game had be designed a little smaller with fewer features intended for launch. Then they would not have had to spend six months of more trying to patch fixes rather than concentrate on some of the other features they had planned.
The same goes for Tabula Rasa. And also Age of Conan.
These games are simply too big for their planned launches. They need to trimmed down very early on in the process. Smaller worlds with fewer features should be the design theory. Developers can add anything they want to their games after launch, but if they are forced to spend their first six months fixing the entire game then they can't get to the "new" features in a timely manner.
Sure, game launches will likely always have hiccups, but the biggest problems should be server oriented (which can be fixed within days) rather than due to buggy game code that takes months to work out.