That kind of innovation (a) is difficult and (b) requires precisely configured conditions. Unfortunately the lesson that corporations learned from the golden age of these labs was that though they produced amazing things, they didn't provide an adequate return on the investment. Now that everybody has learned that lesson, there are not many such labs left and the ones that are thriving operate much like either product incubators (Microsoft) or agencies (PARC).
Another problem is that there are not many properly educated people to establish and run labs like this thanks to our education system falling apart. But you won't hear anybody talking about that. Things will just get worse and worse.