This has been my introduction to Robb's writings and I must say, I'm impressed. From the blog
Iraq's state is hollow -- it exists but without any meaningful functionality.... A hollow state doesn't provide the first two layers, which by extension makes the layers above it inconsequential (contrast this to the US strategy for Iraq). These services, to the extent that they do exist, are being supplied by black markets (from gasoline to the 2 MW ad-hoc electricity network in Baghdad) and neighborhood militias. In order to get these services, the population has replaced any residual loyalty to the state with primary loyalties (to tribe, family, mosque, religion, gang, etc.).

Now that this transition to primary loyalties is complete, the feedback loops of ongoing decentralized conflict ... and large scale systems disruption ... will ensure that it stays that way. Also, the alternative transnational market structures of black globalization, driven primarily by the smuggling of hundreds of millions of dollars in bunkered fuel every month, will provide much of the fuel to keep this situation going.

Robb provides a different view of the dissolution in Iraq. The conventional wisdom is a Shia'/Sunni civil war. But Robb both here and in the essay, sees a highly fragmented but efficient set of groups [the open source war cemented by a primary loyalties in a decentralized conflict] operating independently but making marriages of convenience to cooperate when it's in everyone's self-interest. And, from the essay...
Indeed, because the insurgents in Iraq lack a recognizable center of gravity--a leadership structure or an ideology--they are nearly immune to the application of conventional military force.
I respect Robb's arguments but still think the primary loyalty is religious sectarianism. Recent events following the destruction of the sacred mosque
The nature of the Iraq war has been changing since at least the late autumn, when political friction between Sunni and Shiite Arabs rose even as U.S. troops began carrying out a long-term plan to decrease their street presence. But the killings accelerated after the bombing on Feb. 22 of a revered Shiite shrine.... ... that Sunni-Shiite tension sharpened when insurgents destroyed the golden dome of the Askariya shrine in Samarra in February and vengeful Shiite militiamen rampaged through Baghdad and other cities.
There's no doubt that a free-for-all is still in place with little intra-sect discipline. There are a number of Shiite political parties who must agree to unify their political block in any new government. One is led by Moqtada al-Sadr

who seems to be operating independently of the Grand Ayatollah Ali al-Sistani or incumbent prime minister, Ibrahim al-Jaafari. From the NY Times Iraq's Premier Is Asked to Quit as Shiites Split

Any dispute between the Shiite bloc's two biggest factions -- Mr. [Abdul-Aziz] al-Hakim's party, the Supreme Council for the Islamic Revolution in Iraq, and the party led by Mr. Sadr -- carries with it the possibility of armed violence. The factions are longtime rivals, have backing from Iran and operate militias with members in the Iraqi security forces. Their militias fought street battles last August throughout Baghdad and the south, even hijacking double-decker buses to storm office buildings.
Robb's view is true as far as it goes but I don't think its the end of the story. The greater conflict with the Sunnis could escalate, which would unite all now disparate groups along sectarian lines.
"open source war" or "open source insurgency" is something that the early blog/rss crowd gets.  Parallel innovation without a center has been the tech recent theme, and now Robb (and others) are positioned to identify it in war.

I think we've said in the past that the game changes when insurgents can surf the web (including Robb's site) for the best ideas.

And the insurgents in Iraq have net connections.

There are the different factions in Iraq, Shia/Suni/Kurdish, trying to fill in vacuums in power.

As is probably not widely know, there are also different types of "insurgents", or "terrorists", whatever. On one hand we have the real terrorists, sometimes foreign, who fight on idealistic grounds, or with the sole purpose to destabilize Iraq, aka Al Qeada. A substantial part of the resistance on the other hand, are certainly ordinairy Iraqi's who want the foreign occupants out. These people can best be described as freedom fighters.

These two groups don't cooperate, they despise eachother.

You are talking past what I said.  It is not necessary for people to "cooperate" in order to create distributed centers of innovation in a network connected world.  All the players need is info - reports of the techniques and sucess/faliure of those techniques in other venues.  That is sufficient, without cooperation, for continued innovation.
This with open source warfare brings new meaning to the "cathedral and bazaar" anology of OSes. Time will tell if open source governance can work or not. Asymetrical warfare is by definition open source and centralised opponents in a war. Our own Revolutionary War was open source v. centralised. (colonial insurgency v. Brits)

The open source side has the advantage of cells and groups sharing techniques analogous to Linux coders sharing code to make the next kernel. The Internet facilitates an insurgency just as it facilitates Linux development. Plus, the open source side always have people join and "retire" at will, making intelligence by the centralised side impossible.

The drug war is just as unwinnable as the police have an open source opponent in the form of dealers who come and go as well as manufacturers and importers who come and go. Neutralise any one, and like Whack-A-Mole, a new one emerges to fill the vacuum and supply the market with the not-so-goods.

History is loaded with cases of insurgencies fending off superpowers. Gangs v. cops, the drug war, our Revolution, the French Revolution, etc. And it's found in fiction like Star Wars with an insurgency in personal planes (their flying cars?) v. a Deathstar.

Another advantage of an insurgency is that they can innovate to seemingly no end. The IEDs are a prime example, along with smuggling methods now surely being adopted from drug smugglers. And the biggese advantage is that centralised opponents NEVER learn. Irony of ironies: Now we Americans are the "redcoats"!

This with insurgencies brings up a nightmare scenario for the UN. In America there are 200 million guns in the hands of citizens, along with people who can make IEDs. America collapses, and an Iraq-like situation ensues. (only way bigger) When the peacekeepers come, well you can imagine now with Iraq being the precedent.