I would split theoretical ability and statistical ability into separate skillsets. Having the understanding in the field and the ability to posit reasonable hy

I would split theoretical ability and statistical ability into separate skillsets. Having the understanding in the field and the ability to posit reasonable hypotheses to test is very different (at least in most fields!) from having the statistical knowledge and ability to properly generate results. And I think larger labs do frequently separate out a few of these. I'd love to have a better understanding of how industry labs differ from academic labs in that respect.