> end of the day, hydrogen is always going to cost more than the energy source that was used to produce it.
What your failing to account for is a.) the fact that Hydrogen (and Oxygen) used with fuel cells is effectively a battery, and b.) all batteries use more energy to charge than the amount of retractable energy, c.) we accept these energy losses (costs) in the battery charging process because we need batteries for most mobile applications (not subways & similar though, these are connected to the grid).
H2 and other batteries (including oil itself) are a means of making energy portable so that you don't have to plug into the grid. If we could pluge everything into the grid, then yes you're correct, H2 energy storage would only be useful for back up grid or grid buffer energy storage.
For stationary applications, it doesn't make much sense to have batteries (H2 or Electric, or otherwise) since, as you state, you can just connect to the electric grid and avoid battery charging losses (unless your worried about grid instanility). However, for mobile applications (cars, ships, planes, etc), now you need to consider how to store energy off of the grid in a manner that is COMPACT and LIGHT, because these factors greatly effect vehicle transport efficiency.
Oild is still king in terms of volumetric energy desnity and specific (mass) energy desnity. An energy capacity equivalent Li-ion battery is still relatively much heavier and takes up more space (meaning your vehicle will be less energy efficient). H2 storage has the potential to exceed oil in terms of volumetric/specific energy density, but it's not there just yet.*
*Note that, the other component in the energy storage race isn't just the storage media, but the energy conversion efficiency of devices that use these media. For example, most common fuel cells are roughly 50% efficient in terms of the energy usage, but if you can make a fuel cell 90% efficient, then that makes the use of H2 as a battery, much more competitive. The same is true about oil burnjng engine efficiencies. Electric batteries and motors are very efficient at converting energy, however, electric batteries are currently much less energy dense than H2 or oil systems.
Oil is effectively a battery that has been charged naturally by the sun (as with most mined chemical energy storage). If you create oil in a lab, you are storing the reaction energy for later use, just like charging any battery.
When you charge a battery, because batteries follow the laws of thermodynamics, you always invest more energy to charge that you get out of them. Mined oil is cheap energy because it's already "charged". However, if you were to capture the waste emissions from the atmosphere and recombine them, the energy lost in the recombination would be far greater than that which you can extract when you burn it again.
If someone had a way to economically capture oild emissions, then recombine the into oil - and if this process is cheaper that liberating H2 from water.