> The stats of breakthrough cases are skewed due to the CDCs methodology of tracking covid infections among the vaccinated.
I guess, but the CDC isn't the only reliable source of information, and from everything else we see (international studies and natural experiments from things like hospitals testing all arrivals for all causes), the vaccine still has something like a 50-90% reduction in infection rate, depending on the particular variant and study. And beyond that, a 90%+ reduction in sever cases.
> There have been cases of covid infections among enclosed vaccinated groups
Yes, as the CDC website says, in bold "Vaccine breakthrough cases are expected". What we know is that the vaccines appear to be highly effective at preventing severe cases, and slightly less effective (though still pretty darn good) at preventing infection whatsoever.
Bayes rule would suggest that as the majority of people become vaccinated, we'll begin to see a larger and larger share of infections coming from vaccinated individuals. But, we'll also see fewer infections (and fewer deaths!) overall.
While we don't know precisely how delta compares to the initial variant, we do know its more infectious, and still quite deadly (from its impact in India). So we do know enough to draw the conclusion that Delta is dangerous, almost certainly more dangerous than the older variants, and we also have data to support the claim that vaccines are more effective immunity against delta than natural infection. That's enough information, or if it isn't, what else would convince you?