I've seen this repeated often, and I've never seen any convincing evidence to support it. The studies I have seen always focus on "task-based" roles, where productivity is easily measured and everything is very well-defined. Think call center employees, stuff like that, where they're simply measuring call volumes.
But execs are clearly more worried about the general "disattachment" people have with their job when they're working remotely, and they're worried about reductions in innovation and collaborative idea generation. Those kinds of concerns are much more difficult to measure in a short term study, but I think they're very valid.
Since those things are difficult to measure, I'm not saying there is good evidence one way or the other on remote work's effect on those things. I am saying I think it's disingenuous and just a bad argument to say remote work has no impact on productivity.