Also: stored-energy locomotion. Fully-powered walking is expensive.
1. Run silently or near silently. They're generating electricity via redox reactions, not combusting.
2. Have effectively no moving parts. Again, no noise or vibration.
3. Product electricity directly. For powering electronics, this means no secondary generation required.
4. Utilise extremely expensive catalysts. The reason you don't see fuel-cell automobiles (excluding very small numbers of test vehicles) is that the engines cost, literally, on the order of $1 million, as opposed to a few hundred for a typical ICE. This price premium has proven difficult to overcome.
5. Operate better delivering fairly steady-state power, from what I understand. A tremendous benefit of ICEs is that they can scale output in a second or few seconds, with very little throttle lag. Yes, straight-up electric motors are even better at this, but straight-up batteries tend to not store much power.
A fuel cell is essentially a battery whose electrolytes you fill rather than recharge or replace. This offers the advantages of fuel-based systems (high energy capacity) and batteries (direct electrical output). Unfortunately, it introduces the disadvantages of fuel cells (cost, power delivery profile).
I was thinking of hydrocarbon fuel cells, to be honest.
But if you're trading an abundant, variable energy source (say, peak solar which is otherwise wasted) for a scarce but valuable and useful store, that's a net win.
Note that whether you're formulating H2 or hydrocarbons, the energy cost is about the same (the carbon sequestration from seawater mentioned above is low energy cost), though it requires fairly substantial capital investments.
One other alternative is the Li-Ion batteries that we already have.
In any case, I am just reacting to the hype we sometimes see about an `H2 energy future', as if hydrogen was a source of energy.