People are working on catalysts that will improve the efficiency of water electrolysis, but they are still mostly theoretical metamaterials that don't even have working quantities to test. Even with catalysts, you can't get around the conservation of energy, it will always take at least as much energy to separate hydrogen and oxygen as the energy you get out of putting it back together in a fuel cell. The storage of hydrogen under high pressure is a bad idea. The containers must be extremely strong and with no flaws and be able to be produced in large numbers while keeping quality at NASA rocket propellent storage levels, and we know even they blow up occasionally. The smaller the molecule of gas, the harder it is to create reliable seals. Hydrogen is only slightly larger than helium (hydrogen exists as a diatomic molecule of two hydrogen atoms while the slightly large atom of helium exists singly) and will leak through just about any standard seal, especially at high pressures. This means that all the transfers of hydrogen will require some sort of closed transfer mechanism that prevents the car owner from causing any potential damage to these seals. We already have problems with LPG (liquified methane) verses Propane/Butane. LPG requires very high pressure to store compared to propane and butane, so currently LPG is almost exclusively used by commercial vehicles where the larger heavier tanks can be accommodated and fueling can be handled under controlled safety conditions and still many believe the safety of using LPG is not quite where it needs to be. Hydrogen will not liquify at normal temperatures, meaning the only way to store very much is under extremely high pressure, far more that LPG. Most serious engineers believe the only practical and safe way to store hydrogen is to develop storage systems that absorb the hydrogen in another material under lower pressure. This is how acetylene is stored, by dissolving it in acetone (it will explode if just stored under high pressure). But again we are no where near finding a suitable candidate material for hydrogen.
Since in both batteries and fuel cells we are effectively using energy from the grid and storing it in some medium to be used latter, the question then is how much is lost along the way before it actually drives the wheels of the vehicles. I'm not sure of the efficiency of either system, but I do know that it appears there are far more paths to improving battery efficiencies currently being explored than fuel cells. Super capacitor technologies hold the promise of approaching 100% efficiencies and they are now approaching energy capacities per unit volume that make them competitive with chemical batteries in some applications and there is good reason to believe capacities will continue to rise and eventually exceed that of chemical batteries.
The best application for fuel cells is simply to not use atomic hydrogen. Fuel cells technology exists that can use simple hydrocarbons and atmospheric oxygen to produce electricity directly. They are more efficient than internal combustion engines and produce only pure CO2 and H2O. There is also no reason to use petroleum based hydrocarbons. They work just as well with alcohol, and although current corn based alcohol production is a disgrace, other more ecologically sound methods exist and have the potential for significant improvement in production. So this path provides the potential of approaching a net zero carbon footprint. Why is this not being pushed?? Because of policies put in place by Bush's "hydrogen initiative" restricting research grants to hydrogen only and by environmentalist sentiment among the developers of green technology thinking hydrogen is greener that any hydrocarbon. Small scale versions of these fuel cells are already being developed with the intended market of powering portable devices like laptops and cell phones which will run far longer than any current battery technology and are recharged simply by injecting a couple of ounces of alcohol.
= new reply since forum marked as read