Reinforcement learning is a helluva drug. I'm sure by now Waymos can time yellows in SF to within a nanosecond, whereas humans will only ever drive through so many yellows will never get that much training data.
It's not just collecting the information; it's the ability to spread it.