With galvanic cells, we assume that the redox reaction is kinetically inhibited so that equilibrium takes a long time to reach, so we can make a good measurement of the potential difference V. I thought fuel cells were the same originally, except that we recycle in reactants and products to make sure that there's enough of each to maintain a high potential V, so that we produce power (equal to IV where I is the current). But now I came across the idea of using a catalyst with the fuel cell. That just doesn't make sense to me. The more the reaction occurs, the closer the system will get to equilibrium where V=0 and thus the power output is 0, so why would we want that? OK, so we are cycling in new reactants anyway, so the reaction will never be at equilibrium - but still, why would we want to catalyse it? What's the benefit in that, when the power output is based specifically on potential difference (which is a function of how much of the reaction is still left to go at any given moment in time)?