In some ways it's equivalent to solving any other seemingly esoteric theoretical problem that people seem to like solving (and some people get PhDs for doing so!). He took a formal system (a particular simplified model of city simulation), and produced a constructive proof of some properties of the system. Of course, the SimCity simulation parameters are unrealistic in a lot of ways, but it doesn't seem inherently any more absurd than the people who get PhDs for game-theory theses that also rely on absurd parameters in their problem setup.