A fairer comparison would have been [ total number of reported crashes ] / [ total number of cars on road ].
Comparing it with sales growth will show poor performance for older brands which have a lot of cars on the road vs newer brands that are just release cars now.
I suspect the real difference is that Tesla drivers cede far more autonomy to the car than other drivers. When you look up from your book and see that you are 20 feet from a collision at 65mph, it's too late to do anything about it.
Perhaps dropping $10k or more on something called “full self driving” makes said drivers believe that the car can fully self drive by itself?
For me it would be having used "full self driving" mode, diligently, for X years or X miles without any issues.