An increasing number of Kaddafi supporters, of all sorts, are reconsidering their allegiance. In short, there are far more people going over to the rebels than the other way around. The basic situation is that Kaddafi is cut off from the world and being pounded by NATO aircraft, as rebel ground forces keep pressing in. NATO lawyers are talking war crimes trials for Kaddafi, and many of those close to him don't want to be dragged into that. Another incentive to physically get away from Kaddafi is his threat to use his supporters as human shields outside buildings and military bases threatened by NATO air attack.
Most Kaddafi supporters are in Tripoli and people there know that the NATO air strikes are very precise and cause few civilian casualties. They know that Kaddafi's Information War campaign to stop the bombing, by deceiving the world into believing that NATO is killing large number of civilians, will fail. The rebels smell victory, and will not agree to a ceasefire or truce. So the defections will continue and increase. And so will the bombing.
Strategypage also discusses how battling the Berbers south of Tripoli is diverting loyalist resources from the battle against the main rebel forces.
I guess I don't question the weakness of the loyalist side--they have lost control over large parts of the country. And I don't doubt the capabilities of NATO air forces to precisely bomb targets. So it may very well be that at some point loyalists throw Khaddafi under the bus to save themselves.
But I still think that some event that saves Khaddafi is more likely to happen before the weakness of the Libyan regime results in a collapse. Powerful countries don't want Khaddafi's regime to fall (bad precedent there); other countries and organizations not powerful also don't want Khaddafi to fall and might prop him up just enough to survive; the UN Secretary General wants a ceasefire for humanitarian reasons; some NATO countries are not on board an endless and expensive war; greed will give Khaddafi an opening to get some funds flowing back to his regime; and collateral damage could yet spark outrage in world public opinion to undermine the will of coalition partners.
If current trends continue, NATO could indeed win this war. For the Iraq War, that was pretty much my judgment, too, worrying mostly about whether other trends to quit would take effect before we reached a tipping point in the war where success was too obvious to ignore and too far advanced to be defeated by growing trends to accept defeat.
So in this war, too, I see conflicting trends. Strategypage's trends are true enough. But they aren't the only trends involved in determining the outcome. And how do we know that the key loyalists around Khaddafi have an accurate view of that pro-NATO trend? If the loyalists believe they are winning, they'll keep fighting--and provide more time for the counter-trends to bear fruit.
I hope we win this war--even if that outcome seems to contradict my land power-centric view of how wars are won. But wars are often close run things with one side coming close to absolute defeat before recovering and winning. We surely are making slow gains in this war. Yet if NATO gets close to winning but in the end falls short, we won't get partial credit. We'll simply lose the war.