It sounds like you might have misread that sentence, perhaps reading "better reliability" rather than "better reliability information." At least I don't think I said anything about continuous improvement or the redesign of parts.
As for the other part of your statement, people often think this--when they haven't taken the time to consider TrueDelta's unique research design.
Specifically, you suggest that only people with problems to complain about will sign up and participate. But unlike other surveys, people cannot report past problems on this one unless they are very recent. The survey always only covers the previous month. Instead, owners are asked about their experience AFTER signing up, with a response needed four times a year.
While a random sample with 100% participation would be better, such a thing also never occurs. This is the next best thing.
The results bear this out. Within the panel overall, a large majority of members have yet to report a repair. In any given month, about one in ten members reports a repair trip. To me, this seems about right.
I also suspect that the results are fresher than you're aware. They're updated quarterly, and promptly. The current results reflect experiences as recently as the end of September. And in February we'll know about experiences through the end of this year.
Thanks for helping me stay on my toes