As someone in logistics, I'm more interested in the actual performance.
Amazon's operations run at scale that is usefully simulative of real delivery-truck operations, more telling than the performative 'tech demonstrations' of other companies (eg a truck or two that they trot out for pictures when the Sustainability C-suite is giving a speech) where it's impossible to discern if the trucks are providing a value/performance that means EVs are *actually* interesting for businesses.
The comments to the OP referenced article are few, but the one is positive:
"The Amazon Rivian vans are great. Quiet in the neighborhood, and the drivers I have spoken with love them. They also carry more packages than their older trucks. "
+ Drivers love them is a huge plus. They could make all the economic sense in the world but on a practical level, if the drivers hated them they're going to find reasons to make them fail.
+ More packages than other trucks is also huge.
In short, this is promising. Local-region delivery vans that can charge nightly, deliver during the days, lots of start/stops (ie the absolute WORST performance envelope for ICE in terms of efficiency, wear, and pollution!) is a great place to see where EVs can leverage their strengths.
So far, so good. But let's be honest: while the costs for this are large in absolute numbers, Amazon profit was $30bn last year. In terms of what they spend on shipping logistics, $200 million could still be a performative, boutique, tech demonstrator for them.
One question I would also like to know is regarding these vehicles being custom-built for Amazon purposes. I am curious if somehow the EV 'frame' is more amenable to easier/cheaper/more varied internally custom body builds than that of an ICE: that could be a compelling plus in favor of EVs as well? Would Amazon's mentioned benefits - driver preference and better capacity for the kinds of loads they handle - have been available in a custom-built ICE vehicle? If not, why not?