Rupert has mentioned the 40% energy cost of compressing H2, but I keep thinking that if you have enough excess capacity of solar/wind, you don't actually have to care that much about efficiency. You can use the excess capacity to generate and compress H2 when there is a lot of sun/wind and not enough demand (as happens now in Germany and California).
It reminds me of the idea of a Bitcoin battery. Bitcoin, obviously, has a huge energy footprint and it is not actually a store of energy in any way. But it can sort of function that way in this sense. If you have a ton of capacity that you cannot control like wind and solar, you have to throw that capacity away when you have demand. But one possible application is that as price drops, it becomes attractive to Bitcoin mining. So they buy the excess, but only the excess capacity because at full market price it isn't worth it. So effectively you turn excess capacity into money and subsidize renewables. That's one possible way out of the current negative environmental effects of BTC mining.
