I have a lot of respect for Col. Lawrence Sellin, as you can tell from my Danger Room piece in August reporting how he got unfortunately shitcanned from ISAF for the crime of bemoaning its bureaucracy in public. His SWJ piece today... I dunno. The argument that Pakistan is the center of gravity of the Afghanistan war is incontrovertible, unless you unmoor the war from its al-Qaeda-centric goals (not a good idea). It's also not particularly controversial. His controversial point, though, is a bit orthogonal to the actual point:
The counterinsurgency approach currently being executed in Afghanistan will not by itself achieve the goals of defeating al-Qaeda and denying the Taliban an ability to overthrow the Afghan government, even within the recently announced expanded time-frame to 2014.
That's true, but. Focusing on counterinsurgency makes less sense here than focusing on war. Sellin is really expressing the generally-accepted point that there's no military solution for a war like Afghanistan. When it comes to evaluating the actual warfighting strategy in place, counterinsurgency theory does a better job of recognizing that insight than its more-kinetic competitors*, something on my mind as I've been working on a piece on reintegration efforts lately. So if we're going to discuss the particular impropriety of COIN for Afghanistan within a framework of the war being a poor tool for Afghanistan, then it's necessary to compare COIN to other warfighting strategies that stand a better chance of mitigating the central flaw. Personally, I'm unconvinced they exist.
In the interest of not taking my own side in the argument: I suppose a rejoinder could be, Well, no, because we're not talking about a hypothetical war, we're talking about what the Afghanistan war actually is, so it's appropriate to focus on the war's central lacunae as a problem counterinsurgency fails to address. Fair point, but! Again, that's not a problem unique to counterinsurgency, it's a problem for the Afghanistan war, full-stop; and since there's a meta-debate about the value, wisdom and propriety of counterinsurgency, it's better not to conflate the two. If you really want to point to the shortcomings of counterinsurgency in Afghanistan, then look no further than Gen. Petraeus' recent emphasis on the kill-capture-bomb aspects of COIN, however much he'd argue that all that stuff is as much component of counterinsurgency as the tea-drinking is. That's not wrong -- see below -- and every counterinsurgency strategy is different, but it does represent an implicit rejection of the High COIN tactics of Gen. McChrystal.
*Points we should all be able to agree on: counterinsurgency is war; war is inescapably violent; counterinsurgency's observation that violence is insufficient for achieving a war's political objectives and sometimes counterproductive to them does not make counterinsurgency a sanitized, "Kinder, Gentler" method of warfare. I should write a manual or something so we can spare a lot of needless lexicographical argument.