It's my sense that innovation in the last few years has radically changed in basic structure and style. This feeling predates the current economic stress, and has nothing to do with the 2008 article published in The Atlantic magazine, "Is Google Making Us Stupid?". (Still, a good case can be made for that proposition. For example, have you observed some store clerks lately struggling to make change on a small purchase?)
It seems that R&D and innovation have been transformed into a linear, almost deterministic process. It's as if we know where we want to go, and it's merely a matter of developing the techniques and overcoming some obstacles to get there. The chip industry's much-promoted roadmap is a clear example of this sort of innovation approach; it serves as both a guide and a straitjacket.
The Internet fosters this sort of targeted thinking. You sort of know what you are looking for (i.e. "metals resistant to X"). You search and soon, in many cases, you have what you need, or at least a good idea. It's all so straightforward, and the Internet and search engines excel at fostering this sort of research. If you know where you are trying to go, it's wonderful.
But if you look back at the realities of game-changing innovation, in most cases it has not come from this sort of directed inquiry. It comes via serendipity, coincidence, fortuitous cross-linking, adapting developments in unrelated fields and hard-to-anticipate synergies.
One innovator develops a strong, thermally-conductive plastic of no immediate application; someone else reads about it in Popular Mechanics and says "Hey, I can use that for packaging ruggedized power supply modules."
One of the virtues of random-walk print browsing, compared to Internet searches, is that it encourages this sort of "This-looks-interesting" inspiration.
There are many well-documented examples of innovation triggered by adapting unrelated developments. James Burke's "Connections" series made this point convincingly and dramatically. Would linear R&D have connected the Galois fields of algebra theory (circa 1820) to Hamming error-correcting codes (1950) and Huffman data-compression encoding (1952), and then to the CCITT G3 fax machine standard (1970s)?
On the other hand, maybe one shouldn't blame the Internet for stifling innovation. My concern focuses on how the nature of true innovation has changed. Is the real problem that we have advanced so far technologically that no one can keep up with or understand advances in other fields, nor has time to do so? Go to any technical conference and you'll see tremendous progress and knowledge, for sure, but mostly within the context of niches within niches. Look at conference papers or journal publications, and you'll see how much there is to know about so many narrow topics.
So maybe it's not the Internet's fault, but our own, that advances have radically changed the innovation game into one that we cannot win the way we once did.