It also fails to mention how the accident rate compares to human drivers.

That may be because Tesla refuses to publish proper data on this, lol.

Yeah, they claim it’s ten times better than a human driver, but none of their analysis methods or data points are available to independent researchers. It’s just marketing.