> that's arguably why the j&j had a much lower rate than Pfizer, as it was tested amongst test pools with different variants circulating.

J&J was tested in populations that didn't have the SA variant, and Pfizer was tested in populations that did (in addition to the other combinations you mentioned). Tests were done worldwide and we had natural experiments due to different tests in different time in different (relatively) isolated populations.

That effect exists, but only covers a small percentage of the difference in efficacy.