Peters is controversial and cannot be considered an encyclopedic authority, but in this instance appears to have hit upon a strategy that is not only more conservative in terms of risk exposure than either the "tar-baby" or the "antiseptic surgery" options, but also nullifies Taliban advantages and NATO's current strategy vulnerabilities.
Excerpts from his premise set-up:
"... our grand ambition to build an ideal Afghanistan dilutes our efforts to strike our mortal enemies, mires our forces in a vain mission civilatrice, and leaves our troops hostage to the whims of venomous regimes."
"Even a return to power of the Taliban—certainly undesirable in human-rights terms—does not mean that September 11, Part Two, then becomes inevi- table. The next terror attack on the West will not be launched from Afghanistan.
Pause to consider how lockstep what passes for analysis in Washington has become. The Taliban’s asymmetric strategy is not to defeat us militarily, but to make Afghani- stan ungovernable. But what if our strategy, instead of seeking to transform the country into a model state, were simply to make it ungovernable for the Taliban? Our chances of success would soar while our costs would plummet. But such a commonsense approach is unthinkable. We think in terms of West- phalian states even where none exist.
We buy into so many unjustified-but- comfortable assumptions that it is bewilder- ing. There is no law, neither our own nor among international statutes, that commands us to rescue every region whence attacks against us originate."
Peters' argument implies that a middle way fits NATO and Afghan interests, and nullifies the interests and strategy of the Taliban.
They have no viable interest in taking back Afghanistan; that could be answered by another war as devastating to them as our 2001-2002 assault, due to their having become concentrated targets. They do have an interest, and great continued viability, in maintaining a destabilized Afghanistan.
The same non-state actor strategy is being pursued by Central America's organized crime-cum-opportunistic insurrectionists, the Maras. Keep the state weak enough and corrupt it enough that it can't or won't eradicate them, and the weak state's porous borders and ineffectual pursuit capacities are their best protection from international efforts at their reduction.
Taliban "shadow government" talk is either our misreading of what is actually occurring, or our adoption, hook, line, and sinker, of one of their most successful information warfare operations. They may indeed foolishly be preparing for a state takeover and the above analysis may be all wrong. But the most likely explanation of their local garrisoning and sharia courts is giving some structure to their terrifying of the population into submission or at least away from NATO collaboration, and their tax collection is simply a means to supplement their narcotics income with lower-risk income that doesn't require border crossings.
Peters' suggestion implies, We simply wait until any such local structure aggregates a concentrated target, then MAGTF it with all four feet totally unexpected one peaceful morning at 0500, and shortly after airlift in combined forces with close air support to cut off all escape routes.
Over time, this makes the Taliban's attempts to take and hold any territory too costly to continue-- just the strategy they are now using against us, with considerable success.
They would end up as just another strain of bandits operating in the margins, and increasingly to their sorrow annoying the Pakistanis, instead of the robust and proud strategic risk they now are as a result of our current strategy which plays into their strengths.