This article brings up a good point but just misses exploring a key part of it.
The point: Leaders are going to have some view about what the odds of success for the surge are and what the costs are. Examples in the article are odds of success 1 in 10 and costs of 700 to 1,000 additional US soldiers killed.
Now, it touches on the other side in that it says that pulling out has a cost too in that "sectarian warfare would probably intensify." We can assume that failing at surge (the 9 in 10 odds speculated about) also has this cost or a similar one.
What's missing: There is not an actual example of what the cost of pulling out/losing is with the requisite "sectarian warfare". Let's make some up: 10,000 Iraqi dead? 50,000? 100,000? 200,000? What about the ultimate consequence to US interests? Is pulling out more likely to cause the US's enemies to make peace or to fight harder in a new location?
My speculations would be:
1. Most US citizens would say that American lives are more valuable than Iraqi lives and so the cost of pullout is perceived to be lower. Thus the desire for pullout. Personally, having lived through the 1970's when millions of people were killed after the US pullout from SE Asia, I'm not down with that. Particularly when US causalty levels in the Iraq War are very low by any definition of war.
2. Pulling out is not likely to cause the US's enemies to stop fighting and make peace. The war will continue in some other form and theater after a pullout.
3. I think the odds of success for the surge are greater than 10% and actually go up the longer the US stays. Perceptions of early withdrawal are a key part of the insurgent strategy. Knowledge that withdrawal is not forthcoming creates an insurmountable problem. The insurgents simply have zero capacity to force the US to leave if the US doesn't want to. Truly, zero capacity.
So, the US should stay and more Iraqis get to live.