Regarding hydrogen from electrolysis, and use of hydrogen as a combustion fuel in general.
No matter how efficient you manage to make electrolysis of water, it is a net loss of energy. In the theoretically best possible case, you will get back perhaps 80-90% (see note 1) of the energy you put into it upon burning the products. In practice, it's rather less than that. That's fine for extracting oxygen from seawater on nuclear submarines, sure, but hardly anybody uses it for industrial hydrogen production (~4% world hydrogen production; see note 2).
Currently, the primary source of hydrogen is the cracking of petroleum. Hence, hydrogen as a fuel is, at best, merely an inefficient use of oil--or, if one is daft enough to electrolyse hydrogen and think it is a good use of electricity, it is an inefficient use of electricity, which was most likely generated by burning some kind of hydrocarbon. This is fine for industrial processes that require hydrogen. It is inappropriate for fuel (though not *as* inappropriate as generating it by electrolysis).
Converting other forms of energy into hydrogen and oxygen for fuel means a loss of energy. Why not skip that step and just use the energy you've got? The only situation in which I can see hydrogen as a combustion fuel being feasible is if somebody discovers a gas field full of the stuff. In that case, it'll go to people who mean to do something useful with it, like generate ammonia through the Haber process.
An aside: I might make a possible exception for nuclear fusion--of course, that works better with deuterium and tritium, unless you're a star. Let's just wait 'til 2030 or so and hope that the descendants of the ITER are juicing up our power grids by then.
1. http://bellona.org/filearchive/fil_Hydrogen_6-2002.pdf
2. http://www.hydrogenemporium.com/iFRAME_hydrogengenerators.htm