Tired conventional wisdom about this unconventional war?

As should be pretty clear by now, I'm not wholly convinced that escalation in Afghanistan is a good idea. Actually, I shouldn't equivocate so much: unless I see some new and compelling argument to convince me otherwise, I am wholly convinced that escalation in Afghanistan will do little to secure the strategic objectives that we should be pursuing, and that the expenditure of resources required to prosecute an expanded war in that country (and region) will be completely out of balance with the potential strategic gains consequent to that escalation.

Now that we've got that out of the way, let me tell you an anti-escalation argument that I do not find convincing at all: that an increased foreign force will agitate the Afghan populace into greater support for the insurgency. We saw the same argument in Iraq, by the consistently-wrong GEN John Abizaid:

Abizaid had advised Joe Collins, the Pentagon's chief civilian peacekeeping expert, that U.S. forces would be an "antibody" in Iraqi society and it would be important to put an Iraqi face on the occupation.

Secretary Gates has repeated this worry with regard to Afghanistan (though lately he suggests that a change in approach to a more population-centric counterinsurgency could mitigate these concerns), noting that the Soviets had 120,000 troops in the country and worrying that a similarly significant intervention could alienate Afghans.

Which is why it's interesting to read Robert Mackey's blog in the New York Times The Lede, entitled "How Many Troops to Secure Afghanistan." He closes with this:

One man who has suggested that more American troops are not the answer is Russia’s ambassador to Afghanistan, Zamir Kabulov, who was a K.G.B. agent in Kabul during the Soviet occupation in the 1980s. Last October Mr. Kabulov told my colleague John Burns that the U.S. had “already repeated all of our mistakes,” and moved on to “making mistakes of their own, ones for which we do not own the copyright.” One of the biggest mistakes the Soviets made, Mr. Kabulov said, was letting the force grow too large. “The more foreign troops you have roaming the country,” he said, “the more the irritative allergy toward them is going to be provoked.”

Allergy, antibody, immune reaction, and so on. It's almost taken for granted that this is true. My question: WHY? What Gates has been saying lately is much more compelling: Afghans are less concerned with how many troops are in-country than what they're doing! (After all, does some guy in Spin Boldak really see a difference between 68K and 110K if the engagement in his village remains the same?) Does the top-level number even matter to anyone other than Taliban propagandists?

More worryingly, this is just yet another example of a disturbing trend in the conversation about this war (and others): assertion disguised as argument. Ok, so you think that foreign troops are an antibody, and that the population will have an auto-immune reaction to displace them? Then show me when that's happened! Explain to me why I should believe this to be true! Don't just keep repeating the same old canards and expect everyone to believe them. Even the Russian ambassador resorts to this: we lost, so it must've been something about the Afghans, something about the numbers, something about the swarthy Oriental's unique and intractable character. It couldn't be the fact that we did a million other things wrong, like killing a whole bunch of civilians or propping up an unpopular, illegitimate government.

So really, why should I buy this "antibody" argument, especially when its main proponent in Iraq was basically proven dead wrong?