You see this again and again whenever anyone asks volunteers for a useful service. Wikipedia is an obvious example, but often people misread that and say "only a few people dominate discussions." It's true that there is an elite that performs the majority of edits, the majority of which are minor changes. But the bulk of the information was added by a large group of (presumably) specialists. This is a great model--get many people to chip when they know something, and rely on some cadres to edit and present the whole thing (the problem is that the cadres have the political control, and keep trying to keep the amateurs out).
The strange thing is that people are willing to accept tiny wages for working on a whole range of things when presented with a useful platform. At Amazon's Mechanical Turk, you present a task (anything you like, generally menial internet labor) for people to pick up and do. Wages for these tasks are extremely low, on the order of pennies. Yet the workers are not (as far as I can figure) largely from parts of the world where this is not entirely worthless--they just find it more stimulating than fooling around online.
In many ways, these trends make sense. People have plenty of leisure time, and we're seeing a decades-long trend in which the cognitive demands of society, and the cognitive and computational skills, of people are growing. This is the sort of stuff that people genuinely like to do, and the internet enables coordination like we haven't seen since the factory system.
At the other extreme is a notion that such real-time, real-person offered solutions can also dominate. Twitter enthusiasts, for instance, point out the wisdom of the Twitter crowd in offering information on the Hudson River plane and the Mumbai terror attacks. Slate shows how these kind of micro-blogging solutions are often suboptimal for many of our searches--but other socially-generated solutions do often fill the gap.
I realize I'm really late in talking about the 'micro' output method (Twitter, Facebook updates) and the social aggregation (the two are different, though related), and I can't even say how big these will get. All I can say is that they seem to be diverging along two paths--on one hand, expert, specific commentary increasingly follows a power-law dynamic with the best pundits reaching wider audiences, and the best agglomerations of specific talent gaining power. This follows a more general trend online for certain powerful 'portals' to dominate viewership (Google, Amazon, etc). One the other hand, local networks are more important than ever for dominating social lives, purchasing decisions, tastes, and for coordinating actions. The middle seems to be filled with Google and other trusted information intermediaries (NYT, etc.).
An important aspect of internet economies I haven't seen mentioned, however, is how there's virtually no correlation between the welfare a given platform offers and its financial rate of return. Tinyurl is going of out business, apparently, as are a host of other interesting fledging internet start-ups. Ad revenue still doesn't make money online, so you can have crucial systems fail on the basis of profits. That so much of the internet infrastructure doesn't fall apart despite zero accounting (and very negative economic) profits (the torrent sites, most of what Google does, this blog) is a testament to the willingness of millions of people to do things because they liked doing them, not because it necessarily made them rich.
And in that sense, the internet is a very Communist economy. We have spent the dominant part of our evolutionary history in environments that prize our ability to form reciprocal connections with close family and kin; it's not entirely inconceivable that we can act in ways that further our more aspirational goals. Sure, the Soviet Union failed--that was a failure of economic planning in generating an incentive-compatible system of production in which it was in people's own interest to pursue the collective goals. The self-forming character of the internet has, in many ways, generated exactly such a system, which does in fact work.
Current crisis aside, we're headed for a world in which around ten percent of the population is necessarily to produce all the food and material goods we care for. Figuring out how to organize and structure the remaining ninety percent of the economy in the services industry is the key issue for future economies, and it is one that will increasingly rely on relationships and information fueled by the internet. And our financial system is failing us on delievering the platform of the future due to its concern for prices and profits. Well, on the internet, all prices head to zero and no one makes profits--though everybody makes a living. We're not getting anywhere near the level of innovation and growth we could be getting--and Wall Street will never fund it. Either wealthy people start funding this, or we all take some time out of our day to build the new Internet. Even as we don't understand the dynamic nature of current economies or social networks, we're about to create vastly more interesting ones.