This also ignores the realities of renewable energy. The first thing to point out is that renewable energy is terribly inefficient. Solar panels are only 20% efficient and have terrible load factors. And even then, you'll lose most of it without energy storage. As it turns out, a hypothetical system with just renewable energy and li-ion batteries is actually wasting a lot of energy. These problems shrink dramatically with hydrogen since you now have a huge amount of available energy storage. In fact, you can't even reach 100% zero emissions without hydrogen, since long periods of low sunlight and wind occur fairly often, requiring a backup energy source.
Long story short, the efficiency argument is just FUD. People who promote it have no idea what the realities of energy production will actually look like. It's just something people say to block innovation or maintain the status quo.
Right now it takes ~53kWh to make 1 kg of h2, which stores ~34 kWh of energy. Right away a fuel cell vehicle operator needs to pay for an additional 20 kWh of energy for every kg of h2 before they can move an inch. Toyota Mirai consumes about 0.8 kg/100 km, so those 53 kWh of energy move the car about 125 km. That's 42 kWh/100km. My Model 3 SR+ recently consumed 35 kWh to travel 258 km or about 14 kWh/100km. The Mirai takes 3 times as much energy to move as I used on a recent trip. That efficiency gap is a gulf; no matter how cheap creating and distributing hydrogen gets it will always use a lot more power than just putting that power in a battery. I agree hydrogen has a place in the economy, but only where batteries are too large/heavy for the application or the grid is too far away.
Another thing is that you cannot just make up a number and say this is the only possible outcome. Electrolysis efficiency is a moving target, and depending on circumstances, can be dropped to as low as 39 kWh/kg for water electrolysis and even to 33 kWh for steam electrolysis. Note that there are electrothermal processes gives you effectively free steam (like solar thermal or nuclear), so the latter number is fully possible. In the long run, you have to assume that it will not be that inefficient.
Same is true of BEV efficiency. 14 kWh/100km is pretty low of a figure. Not realistic in normal driving for most people. Not to mention you are comparing a compact car to a much larger car. If people are buying SUV sized BEVs, you won't get those numbers. And no, it is not guaranteed that a BEV will always be more efficient. Fuel cells are continuously getting more efficient and if the goal is have big cars with long ranged batteries, battery weight becomes a real problem. There are scenarios where BEVs will lose in efficiency.
Also, a lot of the upside of hydrogen is that you avoid all of the cost associated with batteries. It take less money and resources to make a hydrogen car compared to a battery car. These need to be accounted for in some way. Finally, one major problem is the huge and unpredictable surplus production of renewable energy. There are reasons why we see zero and negatively priced electricity on the grid. There is no realistic way of capturing it all using batteries. But you can capture most of it via hydrogen. That effectively gives you free hydrogen, and is functionally a way to reduce waste.
Ultimately, hydrogen production just continues the logic of wind and solar economics. Neither wind nor solar are particularly efficient, but it doesn't matter because you are using effectively infinite resources. It cost nearly nothing to keep renewable energy farms going. And likewise, it will cost nearly nothing to use that power to drive water electrolysis. The costs will trend towards zero.
