Not that classic Silicon Valley necessarily got the balance right, either: Xerox PARC is the canonical example of a place that was closer to 95% R&D, and 5% shipping profitable products based on that tech. But it seems at least possible that many companies are now swung towards the opposite side: 5% on inventing new tech, and 95% focus on productifying/monetizing existing tech by making nicely polished implementations, pushing it into new market segments, etc.
That might also be fine so long as there's a bunch of unexploited tech lying out there, waiting for someone to turn it into profitable products, but who's going to keep replenishing those stocks of unexploited tech? Will that kind of R&D come exclusively from academic research and a few big industrial labs like MSR?
Basic tech is still there. I think it's hard to know where to look, because like I said elsewhere, things that significantly change our lives in the future start off looking like toys, and not all toys end up significantly changing our lives.
It only seems like we have no basic tech because we're either not looking there, or we're writing it off.
I know facebook and twitter regularly release open source software for infrastructure. Thrift from facebook (and I think Cassandra). And FlockDB from Twitter, if I'm not mistaken. Every other startup's boats rise due to these companies dealing with silly things.