Certainly we end up with dominant companies in respective markets. But how this translates into market power is less clear to me. Dominance does not automatically equal market power, nor is market power dependent on dominance. AOL, for example, was unable to exercise significant market power even when it was utterly dominant in dial up. By contrast, Toys R Us, as a critical buyer in the early 1990s, was able to exercise market power vis-a-vis box stores in the early 1990s in the toy market. WalMart is in a similar position today, despite controlling considerably less than 30% of total retail sales. (30% being the old test for market dominance, still used in the EU but long abandoned here in the U.S.)
Part of the complication is that information markets are notoriously complex and fluid. How does Google make money? "Ads," people say. But this dramatically understates the complexity of Google's reach, and the barriers to competing firms using the same model. Google can make money doing what it does only because it has so many transactions per day. Entering firms cannot even begin to approach the number of necessary transactions for profitability, making them less attractive, further injuring their chances.
This does not make competition impossible, but it changes the nature of competition. Firms go different routes. Kayak, for example, specializes in travel search. This produces a niche market that allows it to compete. Further, head-to-head competition is possible because the cost of using multiple services is relatively low. Facebook is dominant, but Linked-In competes (as does Live Journal) because it is possible to maintain accounts on multiple social networks for different purposes.
OTOH, the complexity of these markets defies easy analysis under Section 1 or Section 2 of the Sherman Act. What exactly does dominance in search mean with regard to anticompetitive behavior when ads are sold via realtime auction? Is the online world really the Shumpeterian ideal when even the monopolist can be dethroned in a moment because of low switching cost and the fickle tastes of the public? Or is it simply that the anti-competitive tethers are too light to detect as they lead us down the walled-garden paths.
I have, on occasion, compared Google to the toxoplasma gondii. T. Gondii can live in any mammal or bird, but to breed must get into the belly of a cat. What's the best way into a kitteh tummeh? By mousie! So when T. Gondii takes over a mousie brain, it flips various natural mousie reactions. Instead of finding kitteh smell scary, the infected mousie finds it attractive. When it sees a kitteh, instead of being afraid, mousie becomes aggressive. Like the victim of some horrible alien mind control parasite, mousie rushes out in front of kitteh, which promptly devours mousie. The T. Gondii are released in the kitteh gut and do their breeding. If you asked kittehs whether they object to this process, I expect they would say: "We lovez teh mousies! Who cares if buggies breed in our tummeh?"
My feeling is that T. Googlii works much the same way as T. Gondii. It alters our behavior. Google (and other info monopolies) not only make it easier for us to do certain things, it changes our behavior as well. We respond to it and learn from it, as it responds and learns from us. The question is: are we kittehs, or mousies?