Ammonia also plays better with alkaline fuel cells, as there is no CO2 to precipitate carbonates (as there would be with methane as the hydrogen source). You still would have to filter CO2 from the air being fed into the fuel cell, though.
I'm not sure about how bad/dangerous the failure modes are (relative to the anhydrous NH₃), but cryogenic O₂ stores quite easily and the purified O₂ should come out of the (solar-powered) production of NH₃, some even already as a liquid.
