I disagree that "the effects of that portion will not be permanent". The safety level can be set to whatever is desired by governments, since governments control how much liability Waymo has. We haven't seen cars get less safe, we've seen governments force car manufacturers to make them more safe. (As well as institute seatbelt requirements, speeding cameras, etc.) I expect the same to happen with self-driving tech. The benefits to driving more aggressively are also likely to be pretty small to the company - I don't think I've ever been in a Waymo ride that's spent more than a minute waiting for pedestrians. So even if they were twice as aggressive, that's saving 30 seconds per ride. Probably not going to have a huge impact on the bottom line.

Also, if you want to include the speculation that they'll make their cars drive more aggressively, you should also include the speculation that the technology will become better and the driving tech will become even safer than they are now.