I don't think there's a clear winner yet. The devil is in the details, as they say. A good Stirling engine will beat a lousy fuel cell, and a good fuel cell will beat a lousy Stirling engine. Both are in the range of 20-50% when implemented well, but they both have lots of problems.

Fuel cells, for example, typically have a membrane at their core; that membrane tends to get clogged with crud. Stirling engines, on the other hand, are generally heavy (and thus expensive) and need to use exotic materials to handle high temperatures, which they need to operate efficiently.