This isn't met as a bash John Robb post, he is a smart guy with some great ideas, but I think he made several mistakes in this post. John was an Air Force pilot, and then a software technician. He also has a great interest modern war/conflict, and it is important to get outside the box thinkers who didn't get indoctrinated in formal military schools to weigh in. He has even successfully borrowed terminology from the software industry (such as open source) to describe unconventional warfare concepts in an increasingly Globalized (new term, I'll see if it catches) world. However, he errs when he assumes that if he gives it a new name it is a new concept and he is the expert who discovered it. To put it in perspective, self promotion is part of the business of selling books, so looking past that at the issue:

The issue at hand is developing militias to solve our problems, to include the situation in Iraq. As almost all readers of this forum know, we have been engaged with militias (and militia like) organizations since our forefathers first settled on the east coast of America, starting with the manipulation of various Indian tribes to fight one another. The Europeans have been doing it much longer, and it even happened during the rule of the Roman Empire. However, John's version of recent history and his assessment for military requirements seems a little off to me.

“The rapid emergence of these local militias in Anbar came as a surprise to both the Iraqi government and the US military”.
First I don’t recall the emergence of the local militias in Anbar or elsewhere being a surprise. Rather I recall various Iraqi leaders approaching us with the idea asking for permission to form these groups, and asking for our support. First no one in their right mind is going to form an overt militia in Iraq without coalition permission and support unless they're fighting the coalition. Why risk unnecessary blue on blue casualties, and just as importantly why not co-opt the coalition's firepower? The reality in many parts of Iraq is that a lightly armed militia will not be effective against a heavily armed and brutal insurgent force, so the only way they could be effective is with the help of the coalition, meaning a substantial number of conventional forces (in this situation) are still required. If the insurgents were then foolish enough to put together a large enough force to threaten the militia group, then the coalition could pounce on them. It was and is much more complicated than this, but the emergence of these groups was not a surprise, and it was a quite natural reaction to insurgent abuses.

Then he added,
“Most beneficially, it eliminates the need for nation-building, massive conventional troop deployments, and other forms of excess. Some questions remain: can the US manage something this complex or this messy”?
I’m not sure how this eliminates the requirement for nation building? In theory we’re looking for long term success, so we’re probably still looking at something along the lines of clear, hold, and build. The militias help us clear and hold, but the government still needs to build, and that will include some form of demobilization for the militias, all of which will require substantial resources from donor states in resources, to include military manpower. Militias are a temporary fix as he indicated. Can the U.S. manage something this messy? We have before, and one can argue it has probably backfired on us in the long term if you look at the Philippines, Burma, Balkans, Afghanistan, etc., where militias formed the nucleus of latter day insurgent groups or criminal groups, but in the short to mid term they were very effective.

I haven't seen much written on how to manage them more effectively so we don't unintentionally sabotage our long term objectives, but I have some ideas on that, as I'm sure most of you do.