One of those mistakes that people keep making is how long it takes to make a MMO. Schedules get pushed back and the game releases massively bugged anyway. Though I have never made an MMO, my experience with planning tells me how long it will take to make the next one: about as long as it took to make the last comparable one.
All those reasons why your game will be done faster and cheaper than the last one? Lies. If you believe them, you are deluded. Your only hope is to skip some steps by borrowing an existing engine, and even then your new features will make your game take as long as usual.
On average, there is no difference between what people predict as the “realistic” time frame and the “best case scenario” time frame. Think about that one for a moment. Your schedule assumes that nothing will go wrong. Where you left slack, you left it only for problems you know you will face, without slack for the unknown, even when you know there will be unknown problems. It won’t help anyway, since your staff is already planning to use up that margin of error like an o-ring.
The fun part is that the more you know, the worse your schedule will be. More detail makes the problem worse. You know that step A will take 10 man-months, step B will take 15 man-months … and no, your perfect schedule is going to fall apart. You will get far more accurate predictions by looking at the last three similar projects and assuming you will take about that long. I have a highly detailed task list, and I use it only to bludgeon work from co-workers. I would never go so far as to expect us to meet the schedule. I mean, I used to, but then you get your first job or college paper, and reality sets in.