Is this a serious question? For one thing, every founder will obviously know whether they are in-person or not, so it's not even possible to do a single-blind study.
I think it's always going to be an emotional/tribal thing. Some people are wired more socially and get value from the bustle and serendipity of an office, some people prefer to be more alone in their comfort zone.
What do you have in mind for a quantitative assessment?
You can make an argument in the other direction too - prove that WFH is better than RTO, or else we should return to the default that modern society has worked in for the massive majority of our history. There is not reliable data on either side - some of the most important things are also incredibly hard to measure. Will company X do better choosing C++ or Rust? Static or dynamic typing? These questions aren't easily answerable with a study.
I'm certain CEOs of FAANG are looking at stats to make these decisions. And what many WFH diehards miss is that your (or my) individual productivity is irrelevant, because all that matters is the output of the company as a whole. I can work really efficiently on the wrong thing and/or not spend time helping others in a way that my overall output for the company, not myself, is worse vs. in-office.
Is this a serious question? For one thing, every founder will obviously know whether they are in-person or not, so it's not even possible to do a single-blind study.