I have no real insight into the engineering and economic optimization scheme for battery-powered locomotives (likely way too early for that). Agreed that there is an important advantage of having weight in the locomotives, but maybe the lead A unit doesn't need as much weight (or power) as the following B units, which could have the batteries and electric traction motors and most of the power. The not-very-good analogy might be the last of the UP GTEL's where the lead A unit had a small diesel-electric system onboard but all the big horsepower was in the trailing unit containing the turbine and the associated generator and traction motors.
As with all-electric vehicles, a major issue is the time it takes to recharge the batteries. An interesting comparison is with a gasoline fired car, taking 5 minutes to fill the 15 gal tank is a power equivalent of ~10 MW (!) (think about the combination of voltage and current to get to 10 MW...). Although ol' Elon is working on it, it still takes time to recharge a Tesla at a supercharging station -- there's also, ultimately, a chemistry limit. So, seems to me the idea of having a buncha B units that can be exchanged for fully charged units at various way-points is one logical way to approach this.
On other advantage of a fully electrical system is the ability to use regenerative braking in place of dynamic braking - so a fraction of the braking energy is usefully used, as opposed to dynamic breaking, which just dumps heat to the local atmosphere.
