None of that really addresses the "is Waymo safer than humans" though?

And the parts that attempt to are misleading (they point at fatality rate as being inconclusive for lack of enough data but ignore that we do have enough data on serious accidents and the results there are very good).