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.