1) The price argument. Hydrogen will generally be more expensive than gasoline; his argument about current prices is like making an argument about gasoline before it was a widely used industrial product.
2) The delivery argument: this doesn't even begin to make sense. Straw man argument where he claims it couldn't be delivered efficiently in the domains he suggests. It is harder to deliver hydrogen than hydrocarbons, but the numbers he quotes are just garbage.
3) Limiting the vehicle storage to compressed gas (muh hindenberg) and cryo. No serious people really talk about this to my knowledge. Metal hydrides (https://en.wikipedia.org/wiki/Hydrogen_storage#Metal_hydride...) are the way hydrogen probably would be transported in a hydrogen economy. It's still less efficient volumetrically than hydrocarbons, but then so is everything else. The argument that it must be as dense as hydrocarbons is just goalposting. People like electric cars just fine, and the batteries they use are 100x less efficient in energy density. At least at one point, the envisioned hydrogen economy was you just retrofit hydride "gas tanks" to existing internal combustion machines. If you can make everything else work, that's a fairly reasonable thing to do. Which is probably why Zubrin never mentions it.
4) Fuel cells; yes, these are very hard to do for a consumer product. They are amazingly cool though, and they should still be invested in. I think this is Zubrin's biggest fear. His goalposting about fuel cell efficiency is silly: so what if they're not more efficient on some measurement than an ideal diesel motor? They don't smell as bad either.
Zubrin is a neocon type who works in the defense industry. For all we know this could have been commissioned by some oil company types, or maybe he's just being cussed. I don't disagree with the over all message though. You'd have to change a lot of major infrastructure to make hydrogen work for a portable energy source.