The far majority of people do grunt work. Only a small group actually does anything both innovative and non-obvious, which is the "use case" presented for these

The far majority of people do grunt work. Only a small group actually does anything both innovative and non-obvious, which is the "use case" presented for these small talks. Most devs who are not part of that small group and find something non-obvious have to fight through many layers of bureaucracy to even get their voice heard, let alone be given time to prove their idea without having to invest their own time. This also doesn't take into account the far majority of those select people have dedicated times where they research potential innovations, and getting into that small group is also difficult. Even supposing FAANG, fortune 100s or whatever top percentile is likely to desire more innovation, the idea spontaneous talks one supposedly only could have in the office significantly impacting the rates of innovation is pretty far-fetched and mostly just an assumption.