So you have no issue with the sample size being so small and exclusively comming from older models of Tesla's s and X lineup . Making a claim about Tesla as a whole today.
Which claim? About the NTSB? Did the NTSB somehow prohibit Tesla from participating in safety testing with only "older models"?
the sample size of the study which the paper was originally based on only had 25 teslas (model s and x from 2016-19 era). Before jumping into the analysis or the methods or techniques used to parse the data. I think it is important to acknowledge that the sample size is hilariously small and represents a very small % of teslas current AP harware software stack.