Roger Cohen deploys his mastery of geopolitics: "Given the nuclear-charged risks, the U.S. must stick with him and maintain aid for now, but with the insistence he move rapidly toward promised elections, restore an independent judiciary, work with Bhutto and get real about quashing the Taliban." But why would Musharraf do any of those things if he knows that our view is that given the nuclear-charges risks we must stick with him? Either we're going to continue granting Musharraf his direct cash transfers or else we're going to make aid conditional. Obviously, this is a difficult policy question. But Cohen's answer: keep giving him the money "but with the insistence" that he do some stuff is no answer at all.
In general, I'd say this is pretty typical of the sort of magical thinking that seems to have infested our foreign policy pundits. How many times have I read a column making an argument like "Iraq is all fucked up for reasons A, B, and C but given the price of failure we have no choice but to close our eyes and hope really hard that A, B, and C vanish for some reason"? It's really foolish, a way of trying to present oneself as wise and knowledgeable about difficult questions without putting anything out there that one can be held accountable for if things don't work out. "No, no," the pundit protests, "I said we needed a policy that works and had no costs this fiasco has nothing to do with me."