Self-selected respondents do tend to have extreme reports, both positive Nd negative. Depending on context one side or the other predominates. On balance the Bloomberg questions seemed likely to generate over reporting of problems. Without much question the expectations of Tesla among enthusiasts have been higher than they tend to be for other manufacturers. In my view it is not possible to even infer that with confidence.
The only parallel I have that seems equivalent was a survey decades ago of California Porsche club members compared with California Porsche owners. We had Polk data for CA then. At that time 964 owners with PCA were far more favorable than were all CA owners, with a larger % of PCA members reporting 4wd problems. Even that was questionable because PCA members were far more likely to engage in motorsports than were others.
without being able to adjust for such variables it is not too easy to judge.
FWIW, for generic products these problems diminish regularly. I trust Consumers Reports for laundry detergent, but not so much for Smartphones, not at all for cars. selection and reporting bias are greatest when the product itself has very strong reactions. Lacking emotional connection there is minimal bias.
