He doesn't cite many examples/sources, but the PARC lab and spin-offs seem to be a huge exception: laser printer, mouse, windows, postscript, ethernet.

Granted, they're not fundamental science, but they were, at least, innovations.