I recently concluded my New Auburn World of Darkness game that started a bit over a year ago. Despite the fact that it lasted more than twice as long as I'd anticipated and I was really ready to move on to a new game (Numenera here I come!), I thoroughly enjoyed it the entire time. I'm going to talk a little about some things I got out of this game.
First off, it's always nice to have a game that actually resolves instead of just "ending". I've done this a few times, but this game might be the one that was most neatly packaged. I don't think I did a "perfect" job (I'm not sure that's possible), but this one was pretty close, if I do say so. I also set up a future sequel in the epilogue, which will be pretty cool - but not for a while (see also: so many games, so little time).
I learned over the course of the game that mystery is difficult to run - at least for my group. You don't want to be too obvious, but you don't want to be too obscure. When the game is about the investigation and putting pieces together, there is a very, VERY, fine line here - and you're going to go back and forth between being heavy-handed and letting the group spin their wheels as they try to figure out what to do next.
By that same token, though, it was fascinating to listen to them speculate when they were into it. So very often they would be right and wrong in an even 50/50 split. I lost track of the number of times I said "Well, that's not wrong...it's just not...quite right either." while talking about their theories after a session. Those were some of my favorite times.
I got pretty good at juggling multiple groups of PCs in different places doing different things. Also, I failed miserably at this from time to time. I think the group was all together less than 10% of the time over the course of this game. The particular flavor and setup just made that unlikely or downright impracticle from a narrative standpoint. It wasn't a "problem", per se, but it did make GMing challenging, and it often meant some percentage of the table was off-screen. Fortunately, this doesn't bother many in my group.
Like many of my games, this one was a total sandbox. Unlike many, however, it was all set in the same place. This might be the first game that really wasn't location driven to any significant degree - so that was a challenging for me. The shifting environment and the various NPCs and mysteries kept things plenty interesting, though, so that helped.
Along that same line, I set out to be very receptive to the players'/PCs' ideas. For most of the game, I just had "the situation", I didn't really have "a solution" to the problem(s) - I figured I'd let them come up with that. This approach worked quite well, and might be the only way I could run a game like this. The group's ultimate "solution" to the big problem in town was one they came up with - not one I'd pre-planned. It had to make sense, of course, but if they'd come up with a different "solution" they could pull off, that would have been just as viable. I really let them drive how this was going to go down, which I try to do in every game, but, in this one, it was taken to a further degree.
I also think I did a pretty good job presenting a game and setting that was both creepy and intriguing - the way World of Darkness should be.
There's probably other things, but those are the big ones coming to mind.