Another interesting piece here, which gets at the complexities involved in realising even the best of intentions in Afghanistan.
The nub of both these articles is clear: the troop increase is only part of the solution. Politics, and in particular the efficacy of Karzai's gov't and its ability to confront corruption and mismanagement, are absolutely critical to bringing some kind of stability to Afghanistan.
Another good read, this one dealing with Karzai's promised anti-corruption drive here.
The terrorist threat to the US, he argues, doesn't need a physical locus of operations. Even if such a locus were to re-emerge in Afghanistan after NATO's departure, Pillar maintains it would constitute no greater threat than that caused by our being in Afghanistan as perceived occupiers. And it would be considerably cheaper to deal with.
So how about this: tell Karzai to shape up (as in purge his administration of figures we know to be grotesquely corrupt and/or hostile to our interests) or we will ship out. And mean it.
Of course, it might be that Karzai is already so weak he cannot deliver the goods -- even if it means his life (or at least a life in comfortable exile -- wearing those great hats). But what then is the point in staying? An endless, bloody battle in support of a weak and corrupt government, with no end game in sight? If this is not exactly a quagmire, then surely it is the mother of all sand traps.