Renewable hydrogen from water is 2-3x as expensive as making it from natural gas†. It simply can't compete at the pump. People won't pay $9/gal equivalent when $3/gal is available right across the street (or across town, for that matter).
† releasing just as much CO2 and fugitive methane, and still being wholly unsustainable
Daniel Nocera has been arguing for years that using solar energy to split water into hydrogen (ie. the "artifical leaf") is the best strategy. Not even nuclear can scale to meet global demand[1]. But solar devices can be mass manufactured and distributed to the point-of-use (leap-frogging the need for massive infrastructure investment).[2]
1. http://web.mit.edu/professional/pdf/DGN_Daedalus.pdf > "Delivering this TW-value with nuclear energy would take the construction of eight thousand new nuclear-power plants. In other words, over the next forty-five years, we would have to construct one new nuclear-power plant every two days."
Yes, we'll need a lot of energy, but note that that figure is for the whole world. I think it's doable. Hell, it's the best alternative so 'doability' doesn't really matter, but I guess we'll see how close we'll get (with nuclear+solar). Note that china alone has 200 nuclear plants in project right now [1].
[1] http://nextbigfuture.com/2015/02/china-could-build-200-nucle...