Yeah, I'm not arguing that commercialization isn't useful; it's the main way theoretically valuable ideas get actually produced. But I think it can be very helpful to innovative industries to have this big stock of theoretically-valuable ideas waiting to be mined for commercialization. Otherwise you either only get incremental advances, or you need big firms with their own R&D groups (like Microsoft Research). You can also have startups based on blue-skies research in some areas (especially pharma), but investors in most areas are pretty wary of funding a startup that has too much basic research in their critical path, due to the high risk.
