Looking back at Tunisia, and forward at Egypt, I think there's an important point that this post almost hits but not quite.
Specifically, I'm fairly sceptical about "Twitter Revolutions" and such - if your revolution has someone else's brand name on it, how revolutionary is it? - but I don't think it's irrelevant.
I'm feeling a little sorry for Evgeny Morozov at the moment. He'd just hacked out a niche as Mr. Grumpy by royal appointment to the blogosphere, when first Wikileaks and then Tunisia and Egypt came sweeping through, and the Tunisian secret police hacked all the Facebook pages in the country, and the Egyptians turned off the Internet, just pulled all the BGP announcements... Sometimes it's not your day.
I do think, though, that there is an important way in which a whole lot of Internet tools contributed to the revolutions. I recently posted on the way in which people can at least for a while function as if they were part of an organisation just because they shared certain assumptions. It's the idea of the imagined community, which can be defined as a group of people who are behaving as if their weak social ties were strong ones. If you want a mental model of this, the revolution happens when enough people change state and start doing this, and it stops again when they revert to pursuing their interests in the normal way. Of course, what happens in between may have changed what those are and how they do it. From a different direction, look at Chris Dillow's post here - it's theoretically irrational to take part in politics, until it's not. The point when it stops being irrational, though, is the point when people stop thinking it's irrational.
In that sense, a lot of the work of starting a revolution is starting a myth. An ironic salute to this was the Egyptian government's decision to turn off the Internet, and later the GSM networks as well. If the value of the Internet really had been as a way to pass on the time and place to assemble, this would have been a serious blow to the movement. But once you're a really angry Egyptian, where else would you protest but Tahrir Square? It wasn't that they needed it for tactical communication, but rather for strategic propaganda. Also, once they took this step, they had also inadvertently demonstrated to the other world media that This Was It. The mainstream media remains very good at bringing its own connectivity, and the main barrier to them covering the news is usually that they don't think something is news. Giving Al-Jazeera and friends - who had been heavily criticised on the Web for being soft on the Egyptian government - a monopoly may have been a really bad idea as it forced them to cover the news or look indistinguishable from Nile TV.
I suspect that a lesson here is that the last thing authoritarian governments will do in future is turn the Internet off. For a start, they will increasingly need to keep it up for economic reasons - the ISP that serves the Egyptian stock exchange and central bank was left alone, and with time I would bet that it would become increasingly porous to information. But much more importantly, this is not a policy that has a great track record. Burma managed it, but started with advantages (not many users, only one network, and a strong position to start with). Iran did far better with its throttle-down-and-spy plan. Even though the Tunisians funnelled all the Facebook accounts in the country into one, controlled by the secret police, it didn't seem to help.
Jamais Casco (via here) asked if you could start a genocide on Twitter - a sensible point, as we know you can do so with the radio, the cinema, television, the newspapers, and (thanks to Serbian turbo-folk) rock'n'roll*. Terrorists tried to start a nuclear war with a spoofed caller-ID. Whether or not you could do that, you can certainly start a mob of quasi-fascist loyalist paramilitaries on QQ. Out of all authoritarian governments, China does best, with strategic trolling and semi-official moderators, which may be more important than direct censorship. Andrew Wilson's Virtual Politics makes the interesting point that Russia in the early Putin years didn't so much censor the Internet, as distribute government talking-points and favours to carefully selected bastards.
Then again, was the greatest success of the wumaodang model the 2004 US presidential election? The best way to fight one myth is perhaps with another. And the best ones are distinguished by the fact they are sometimes called principles. The really depressing consequence of this is that Paul Staines probably has a job for life, although the less depressing corollary is that he gets to herd several hundred idiots yelling about ZaNuLiebour for the term of his natural.
A couple of other interesting links: Charles Bwele makes the point that in much of the world, the so-called new media are more like the first ones. Did you know about the Grozny riots of 1958?
*The world's first genocidal remix is yet to come, but I wouldn't rule it out by any means. All art aspires to the status of music, and just look what people get up to with books.