Bamboo, that seems really, really high for usage, but since I haven't looked it up yet, I'm not sure. A liter of gas is about 1700L (a bit more, actually) of gas vapour, which is what burns. Assuming an high-average (and pretty bad mileage of 10L/100km, counting large trucks and the such, a car gets a kilometer using up 0.1L of liquid gas, or about 170L of vapour. With an average smaller car (the kind with a 2.0L engine, uses about 85L of vapour. Avergae highway speed of about 100km/h, gives us an hour to produce that much fuel, at a rate of about 1.5L per minute. Nt counting precentage of gas burned, etc. That's a lot of H2 to produce, even if we count that H2 is more powerful, and we'll be supplying some O2, but the rest will come from outside the car (this reaction (Fe and Ca), produces, combining the two, 4 moles of H2 per one mole of O2, so we'll need to supply another mole (16g) of O2 ourselves per mole of H2 burned, to burn a perfect 100% (or as close as we can get) of the gas produced.
They normally calculate mileage as stop and go traffic, however, and most cars spend nearly all of their time running at 2000-2500 RPM's, not very far above idle, where use is lower. We will need to supply at higher RPM's as well, and that's where the storage tank/bubbler come into play. Speaking of which, it will need to be fairly large (10L or so, or a combination of smaller containers). Gas has an advantage that it's in liquid form to start with, and stored as such. We work solely (so far) with a gas, and face larger storage problems. Storing the hydrogen in metal hydrides is an option, although a slightly heavy one. It does solve the problem of on-demand hydrogen, though. We can keep the hydrides "charged" all the time, and recharge them while driving, and while stopped. Current hydride systems have a range of about 350 miles of 5 NOS (yes, Fast and the Furious) sized tanks, filled beforehand. If we can at least partially refill the tanks while driving, we should be able to increase the range to a thousand miles, without stopping. When stopped, the hydrogen cells can be made to continue production until the hydride tanks are full, then automatically shut off. Anyone else like this idea?
