Wait, you're getting a decent amount of energy per unit of mass, with zero emissions and complete renewability and... you're complaining of a net loss of energy? Once there's a safe way to lug hydrogen around, just hook up the plant producing it to a hydroelectric powerplant. Voila, no more need for fuel. Hydroelectricity is only a bad option because it can't be deployed everywhere. With hydrogen as a medium (reiteration: once it's safe to use), you can stick HEPPs onto every major river and waterfall, and have very cheap reusable fuel. Also, geothermal. Though I doubt the sanity in the decision to put a hydrogen processing facility anywhere near a geothermally active zone, it's also a valid option.
Fuel is not a problem for a hydrogen production cycle. The real problems are:
a.) Getting the energy. You don't only need to put in a load of energy to split the water apart, but you also need to compress the resulting hydrogen and pump the water and also heat it, to speed up the process and improve yields.
b.) Storage. Hydrogen has a huge energy density
per mass unit. It also has a really low specific density. To get a decent energy density per volume unit, you need to compress it quite a bit. This first of all costs energy, but secondly, it also greatly complicates the design of hydrogen storage tanks and fuel cells, because they too need to cope with such high pressures.
c.) Transport. Hydrogen pipe lines would probably need to operate at high pressure to get a reasonable mass throughput, which means they're essentially kilometer-long bombs. Tanker cars and ships filled with hydrogen are also quite explosive, again due to the high pressure. Unlike oil, which will burn slowly if at all, a broken hydrogen pipeline has a big chance of becoming an instant inferno.
d.) Cars themselves. Again the fuel tanks need to be designed so that they don't rupture in a crash. Any car manufacturer trying to push the prices could make the next Pinto.