OK wrote, "Even if you are dealing with electricity from a solar cell, the efficiency of a battery electric vehicle is not that much better than a fuel cell vehicle, especially as the range of the vehicle increases."

IIRC when we are looking at the amount of noncarbon generation needed to be deployed to meet transportation energy needs, the overall system efficiency of FCV requires 60% more generating infrastructure than does battery electric. That is a HUGE difference and it is the primary reason the DOE has shifted emphasis to battery electric drive. There are also very important value streams related to various storage applications that can be stacked on the battery electric fleet that simply don't exist for H. That not only applies to the batteries in the vehicles being driven, but to the used ones that are rotated out when they eventually drop to only working at 80% of their original capacity.

The decline in energy density is significant for mobile applications, but the residual storage capacity for fixed applications is not at all impacted. In other words when doing a legitimate benefit/cost analysis of EV battery packs you must consider not only transportation value but also their value as grid storage as well as the residual resale value for uses after they are removed from EV service.

It is really a no brainer from the macro view. That's why FERC loves battery EVs.