I couldn't help but look at this slide deck and think about my own experiences with introducing game-like elements into the collection and maintenance of scientific information: Chempedia and Chempedia Lab.
One of the most valuable lessons from Sebastian Deterding's presentation is to start with a specific goal, then create an environment around that goal that makes learning and mastery fun. How do Chempedia and Chempedia Lab do on this point? In my view, not very well. They both contain elements of games to be sure, but in their current form they lack many of the most important aspects of a fun game.
The good news is that the Web is fluid. Having built both sites and run them for several months now, we have a much better idea of what's possible and a more refined sense of those things that don't work well. It might be interesting to combine these insights with a fresh perspective on what makes a good game, and then focus on a problem in chemistry that might lend itself well to a game-like treatment.