I'm not arguing that it's not useful at some level, just that it forms such a tiny fraction of one part of a complicated process that Shep's claim that it can be decisive is laughable, as is his whole conception of counterinsurgency as something that can be solved through military force. Take for instance this:
Which is great, but again, let's look at Iraq: what was more important, the incremental improvements in anti-IED technology or the decision to saturate neighbourhoods with US outposts during the surge? The latter, obviously, and it itself was not nearly as large a factor as the Sunni awakening.It doesn't, not directly. It can be made pointless if you screw up the face-to-face aspect. But done right, it can help: your patrols move more freely and are more visible, making people more secure. You harass random turnip farmers less frequently, making people feel more secure with you than with the guerillas. Guerilla leaders are quickly identified and forced into hiding, making people feel less secure with the guerillas than with you, and so on.
And just quietly, as you yourself acknowledge most of these increments are about force protection for the third party Western force - which isn't a bad thing in itself but isn't doing much to solve the insurgent problem long term. Australian commanders who went to Iraq are pretty open about the fact they used civic aid to tribal leaders as a way of minimalising risk to their men -I'll build you a school if you make sure no-one shoots at us from your area of control. Which is great, cause it meant no body bags, but it didn't do anything to solve Iraq's problems going forward.