I understand it takes 130,000BTUs to strip 100,000BTUs of H from H2O.
That's the issue - electrolysis requires the input of energy to split the water particle, and when Hydrogen is allowed to react with Oxygen, it's an exothermic reaction (releases energy). You obviously need the latter to power the drive battery in FCEVs, and so, unless hydrogen gas is naturally occurring in quantities that are necessary for the demand, it has to be made by splitting water (which takes energy to accomplish). Those factories will invariably use some form of energy to run - be it coal, hydro, nuclear, etc. Based on energy demands, it won't be wind.
I think making ICEs better, along with hybrids (more so range-extended BEVs) are the way of the future. The more I read into HCCI, the more I wonder if perhaps the best thing to do is to make a very, very small generator ICE with HCCI rather than a traction engine since HCCI's main disadvantage is that it doesn't work well at low or high RPMs. You could therefore use the HCCI engine to generate electricity at the optimum efficiency when necessary to recharge the battery.
Or perhaps that's Toyota-Mazda's ideal - to create a better RxBEV? To out-Volt the Volt?