I'm not sure how comparable that data is. It seems to just sum up the times where random Tesla's where in Autopilot mode:
1. Autopilot from my understanding is L3 driving, with the driving needing to be attentive
2. They are not recording disengagements, or even near accidents, only full accidents
3. It seems to be like people would use Autopilot mostly on highways and other places where it works more reliably, so it's not a uniform set of miles
By comparison, here is 21 months of self-driving data from Waymo, with every disengagement and near accidents reported [0]. It's L4 driving with no drivers at the wheel. I'm sure Tesla tests their FSD similarly, and probably tracks similar data, are they releasing that? The other ones you posted says very little.
1. Autopilot from my understanding is L3 driving, with the driving needing to be attentive 2. They are not recording disengagements, or even near accidents, only full accidents 3. It seems to be like people would use Autopilot mostly on highways and other places where it works more reliably, so it's not a uniform set of miles
By comparison, here is 21 months of self-driving data from Waymo, with every disengagement and near accidents reported [0]. It's L4 driving with no drivers at the wheel. I'm sure Tesla tests their FSD similarly, and probably tracks similar data, are they releasing that? The other ones you posted says very little.
[0] https://www.theverge.com/2020/10/30/21538999/waymo-self-driv...