I’d argue that any data collected during the last year (including the data I presented about my current employer, of course) is not actually a very valid measur

I’d argue that any data collected during the last year (including the data I presented about my current employer, of course) is not actually a very valid measurement. I fully suspect all companies (including my employer) that try to support fully remote jobs to re-assess in another 2-3 years time, when life does not have the massive confounding factor of a global pandemic. And perhaps we will find that fully remote employees or teams significantly over- or under-perform individuals who go into the office. So while we don’t really have good data on if remote work is good or bad right now (we certainly know it’s not terrible - most tech businesses thrived last year), companies that decide to not start supporting it at least partially are taking a very real loss (reduced talent pool), for a perceived benefit that may or may not actually be there, that we just can’t measure yet.