The big debate is whether or not to focus on a counter-terror campaign or a counter-insurgency campaign...both of which involve their own respective Gordian knots of unknowns, variables, black swans, and potential outcomes. Nevertheless, much of the debate has focused on concentrating on one of these two camps--either you're for counterinsurgency or for counter-terror.
I'd always hoped that there was something in between these two camps, and I heard it from Tom Ricks just today, during an interview on Fareed Zakaria's GPS (although it's also brought up in CNAS' June publication called "Triage: The Next Twelve Months in Afghanistan", available on Kindle). The CNAS gang proposed using the "oil spot strategy"--standard Galula-based counterinsurgency--which would begin by securing the urban areas of Afghanistan and slowly expanding their reach from these areas. Meanwhile, the counter-terror campaign would be conducted on the peripherals of the ever-expanding oil spots in Afghanistan.
Unfortunately, the down side to many of these "middle ground" solutions, though, is that they seem distressingly similar to what we've been trying to do for the last eight years or so (with an urban elite government in Kabul which seems to barely exist outside of the city limits). Nevertheless, it seems far more attractive than a full-blown counterinsurgency campaign (which, by its very nature, will be messy and slow), coupled with a civilian surge (not going to happen).
The counter-terror crowd, on the other hand, has been sparse on details--they've mainly been "anti-full-blown-COIN" than anything else. That is, until today, when Austin Long at the AfPak channel decided to fill us in (it sounds eerily similar to something posted by Ralph Peters. No no, the counter-terror part, not the killing of American POWs part). Basically, it would consist of a few NATO air bases in Afghanistan, which would also house a few battalions of infantry forces and special operations forces, which would focus on killing or capturing key Taliban and al Qaeda leaders. This minimalist approach would still consist of several thousand troops (and contractors).
The problem I see with this approach, however, is that it will still be very difficult to get the human intelligence necessary to acquire these targets. Without a significant presence from NATO and (especially) Afghan troops, we risk having more incidents like those in Kunduz last month.
Focus: There's no easy way to cut this one. Not to mention, we still don't know the true nature of our enemies in Afghanistan: the Taliban, the taliban (little "T"), and al Qaeda.