The "surge offensive" of last year was largely possible because the informant network had grown to the point where commanders were confident that many Sunni Arab tribes were ready to switch sides.
This is just one reason why the talk of the surge rescuing a failed strategy is all wrong. The war went through many stages that required different approaches. So while we needed the surge for the new circumstances that our prior strategies could not cope with, we did have success under previous circumstances. The surge built on those past successes.
The talk of why didn't we do the surge earlier misses, among other things, that we didn't have the informant network to exploit. One of my worries about the surge early on was that I didn't want more troops milling about in alien terrain just increasing our casualties while increasing expectations at home which would erode our ability to fight faster. An earlier surge would have done exactly this and quite possibly have led to our defeat. Heck, even when I looked at what the surge intended, I worried that it would exhaust our home morale despite being a correct approach militarily. This is because I figured the new approach was more important than additional troops. But the network of spies we developed over time was a key factor in giving our surge a point and giving the new troops good information to act on.
Our success in defeating the terrorists and insurgents in Iraq should indicate how we should prepare future battlefields years before we arrive:
Instead of drawing up nice plans of governing structures and new roads, why don't we spend our pre-war preparation time building up similar databases of leaders and groups in the target nation? This might be a good task for the CIA and other intelligence agencies to focus on, building on those country studies.
Such detailed knowledge of the society and political elites of a potential enemy would be useful for a lot more than just suppressing an insurgency. We could use it to target sanctions, foster a revolt or revolution, or sow dissent and suspicion among the ruling elites.
If we had a database of local actors anywhere near what we have today, we'd have rolled up the insurgency a lot faster and perhaps prevented al Qaeda from effectively invading Iraq and dragging the killing out years longer.
Of course, Congress is working up the nerve to cripple our intelligence agencies as the fears of 9/11 fade:
The U.S. Congress is again, as it did in the 1970s, trying to "rein in" the CIA. Congress wants to outlaw many real, or imagined, techniques that the CIA has employed since September 11, 2001. Much of this effort is political, to placate the many people, and politicians, who now take it as fact (or on faith) that the Islamic terrorist threat was overblown, or that the U.S. response was not commensurate (and itself a form of terrorism) with the threat.
So without people on the ground, we won't be able to study the social battlefield until we are hip-deep in terrorism and insurgency, and we'll have to start from scratch--again. And another president can be unjustly accused of mismanaging a war.
Who knew the power of the purse involved whacking our troops with one?