This raises a myriad of thoughts in my head regarding the relationship between COIN and demographic geography. Basically, what is the extent of the relationship between the security of the population and the dismantlement of the insurgency? There is some dissonance between the two because the logical extreme of asserting that the population itself is the decisive object suggests that we ought to build a McDonald's and Wal-Mart in every neighborhood and hope that somehow the provision of material needs will somehow deflate the political will of the insurgency. I'm not certain that bridge between population security and an insurgency is that steady to cross in full faith. I'm not a historical expert on COIN, so I must ask if there's any example of successful COIN that relied on or emphasized nation-building missions? The successful examples that come to mind immediately (Philliphines, Kenya, and South Africa) all include rough treatment of the locals, which seems to contradict the popular notion that the general population should be treated nicely. That seems to be the case, and please feel free to correct me if my observations are absurd, because those campaigns targeted the specific power structures of the insurgency rather than attempt to appeal to the broader base of the general population So, my thoughts:

The demographics of the insurgency itself -- not simply ideology, but also its ethic, religious, tribal/clan, racial, cultural make-up and what political, economic, and geographic features of the region augment its strength.

1. The combination of these features that shape the insurgent identity (or identities) more or less ensures that it does not represent the population-at-large. If it did, I would question why the insurgency exists in the first place if, in a democratic environment, the state represents the general will of the people (why aren't the insurgents themselves in power?). It suggests in my mind that the democratization of the indigenous state would be an ineffective, perhaps even counter-productive, process, if not irrelevant. To me, it seems not so much a question of human security but of power. The former originates from the latter in my opinion.

2. The insurgent demographics indicate the sources of the movement's strength. As Ken suggested, that strength may not be in locations we would initially assume. In what ways could an agrarian, radical religious, anti-modern insurgency (i.e. the Taliban) find sympathy among the demographics of a major urban center? If there's not much in the way of a power base in the cities, why devote extensive valuable resources to patrolling those cities? Which brings me to point three....

3. How does a military, other than by coercion, establish, generate, and project power? How does a military maneuver through the "demographic terrain" and place an insurgency in a position of most disadvantage? Should, for example, the military promote a favorable version of the indigenous religion in order to counteract the radicalism of the insurgency? Alternatively, should the military adopt a similar version to the insurgency's religion IOT to co-opt their beliefs (much like campaigning politicians do to draw support away from their opponent)? How do we go about shaping the demographic battlefield, if as it's commonly accepted, the people themselves are a part of the terrain, and how do we leverage demographics to our advantage?