Pure science, pure mathematics, pooh. If you want to play with the output you don't get to away from the process and claim to be looking at the data. So we keep going. If individual variation within the cohorts was greater than any intrinsic difference in protection arising from modes of immune system priming it would bin in unpredictable ways between the cohorts. You would see a lot of variation in the studies because you would not be evaluating what you thought you were. So the CDC looked at extant cohort studies in a comprehensive way and that's what they walked away with. So when you say the Texas study, the Danish study, etc I'm thinking there goes Grace chasing noise. In a cohort study, the significance you are talking about is only relevant to the group of people comprising the cohort. Whether it has predictive value outside the cohort is what you need to be looking for.