A simple analogy is if you ordered a navy destroyer crew and my army ordnance company to swap jobs, it end up as a hilariously ineffective reality TV show (Hmm, this might actually lure in viewers...). Something even more extreme like assigning an infantry platoon to peace corps work is going to be even more of an epic fail.
There is another interesting issue that the peace corps is not supposed to accomplish any major long term changes. So if they had a mission there, with the same results, they'd be considered to be doing "great". The S.A. township my cousin served in, remains a morass of poverty and crime and ignorance, maybe a little better than it would otherwise be, but no one has any crazy illusions that his 2 year term or whatever it was would result in the township turning into "leave it to beaver"-ville. On the other hand its considered a military disaster to spend decades basically accomplishing nothing. So you need a perspective change. For a bunch of soldiers doing someone else's job, they're about as successful as the correct set of workers would be.
Interestingly to the best of my knowledge the peace corps still considers hand chalk to be appropriate technology. If you want to do nationbuilding you send in guys with chalkboards and excellent native language skills and maybe some special technical skills (or maybe not).
An interesting thought! One of the most interesting conclusions I drew from this documentary is that the Afghan Military needs both - more technology and more manpower. But the trouble, as shown, is that you need the intelligence to handle so much manpower, the social contracts to retain that manpower and the education to handle that technology.
So what you're saying about the balance might be an interesting point for the US military might work for them, but it doesn't even start to help out with the Afghan forces.
As for going "British-style" - I think that stuff simply doesn't work in the modern, connected world. British colonialism took advantage of isolation and division. Doesn't work so well if the country you invade is reasonably well connected globally.
I don't believe it's feasible to occupy a medium sized country which actively resists using modern militaries. It's certainly not worth it. I'd just focus on keeping a limited presence to rain death from above (like we do in Pakistan, Yemen, etc.) or via limited JSOC strikes (Somalia, Afghanistan before 2007 or so, etc.), and then using armed but civilian organizations to interface with the populace and government.
In a place like Somalia or Afghanistan, foreign organizations should run important infrastructure and logistics services for the local government and population, with the cost subsidized by international donors. i.e. a local Afghan should be able to buy electricity at 0.10/KwH, sell their pomegranates at the world price, and have clean water, Internet, cellphones, etc., without any other focus on building the government (since that's probably hopeless). The organization providing that infrastructure needs self-defense capability, but you don't need much beyond what Blackwater had to do this.
That's what I thought as well - The documentary left a pretty bleak picture of the Afghans ability to build, extend or even maintain basic infrastructure, but it's desperately needed if they want to get anywhere with education. The country needs at least a good dozen generations to heal the most serious wounds so it can find its own interest in the kind of social development that makes it resistant to terrorism taking root over and over again.
Hand-holding throughout that time is what is required. But that's probably the most costly thing, I would guess. Then again, those drones and all that ammunition sure don't come cheap either. But the US has sunk far too much money into that failed path to turn around and do the other one now.
And even then - the problem is that the citizens are already tired of conflict and both sides, the international troops and the taliban, are trying to exploit that. Just giving them great infrastructure might make it easier for the international troops, but it would probably do the same for the taliban.
Whatever the solution, it sure won't be a quick one. My best bet would be to exploit the countries natural resources and push all the money made from that into infrastructure and security. But that's just the kind of "socialism" that nobody in the west wants to pick up.
I'm not sure how well this theory has held up (it's from 2004...), or how much support it has. I think what happened is we got lucky in Iraq in 2007-2010 using the old British strategy of allying with semi-enemies against real enemies (Sons of Iraq and the other Sunni tribes, and non-aggression with the Shia militias), eked out a borderline draw in Iraq while calling it "a victory for COIN", applied the "COIN" model rather than the "British" model to Afghanistan with the surge and all, and have proceeded to lose from 2009-now in Afghanistan.
The US doesn't care that much about the "good" people. They picked the most effective people for fighting the USSR. And out of necessity they needed to use Pakistan's ISI which wanted trained Mujihadeen for their own purposes (asymetric warfare with India) to deliver arms to the "freedom fighters."
[1]http://www.ted.com/talks/thomas_barnett_draws_a_new_map_for_... (Hope the link works, I have somehow managed to break the TED site...)
The appetite for that kind of nation-building has shrunk considerably following the experiences of Iraq and Afghanistan.
That all works well (very well) when you're invading countries. Doesn't work so well when you decide you need to build a new country out of the ruins. Maybe we'd do better if instead of spending billions on flying robot assassins to fight tribal hill-farmers, we simply enlisted more humans to take command.
The only other option I can see is to start running the American empire British-style - they ran very efficient colonies by exploiting existing tribal divisions, usually by recruiting some existing elite as their colonial administrators. Not trying to implement democracy in tribal regions that don't yet have stable institutions or the rule of law would probably also help.