OK, you got me, I screwed up.
What I meant is that a fuel cell is not a heat engine, and thus is not constrained in the same way a heat engine is, as far as max theoretical efficiency in converting the chemical energy of the input fuel and oxygen into useful output electricity.
In a heat engine, you are constrained by the Carnot cycle. The chemical energy of the fuel is used to create heat, and the max theoretical efficiency is governed by the ratio between your high-temperature heat source and your low-temp heat sink. In practical systems, you have a limit to how high a temp your fuel/oxygen mix can generate, and how high a temp your engine can operate at.
In a fuel-cell system, the max theoretical efficiency is 83% versus 58% for a practical combustion engine.
Those are a couple of useful numbers -- thanks!