Waymo doesn't have $2B in R&D. The "Other bets" operating margin was -6B in '22. That would include the army of robotics engineers at Everyday Robots, Deep Mind, Calico and Wing + Waymo. If the cost of running Waymo in '22 was -2-3B in total, most of that would be from operations and not R&D. Perhaps 20% is R&D.
I'd argue the bulk of the R&D cost is behind Waymo and it will likely get lower from here. The sensor hardware is invented, the Driver is pretty awesome.
I agree they're not at 0.1 yet, but I think if you include all roles (excl mgmt) they might be at 0.4-0.5 operational persons per vehicle all inclusive.
Expansion also means more revenue, and the expansion cost is what it is, but that's not R&D. Scale matters a lot, obviously.
In CA alone there was 1.37 million drivers who performed at least one ride or delivery across the DoorDash, Instacart, Lyft, and Uber platforms in California. Gross driver pay, including tips, totaled $1.35 billion in Q3'21 alone. So perhaps 5B for the full year.
If Waymo can get their hands on a 20-50% of that and expand in AZ, TX and a few other sunny states like FL I think they'll do just fine while scaling out to the rest of the US.