.
My belief isn't that a 22 is better for self defense for a person who is willing always to carry something larger.
My belief is that, because a 22 is nearly as good as something larger with respect to probability of stopping an attack, it is a decent choice, and for the person who would otherwise carry nothing, it is an infinitely better choice.
With that said, on to specifics.
the article points out that something as simple as heavy clothing may rule a .22 inadequate.
The article doesn't say that. It says that the test didn't take cover, or clothing, or car doors into account and went on to say that, if you are needing to shoot through car doors, larger calibers are useful. In fact, from tests, many of which you can view on youtube, it is clear that 22's go through clothing.
(A psychological stop, according to your article) a larger weapon is going to be more effective at this.
That is your supposition, but not mine. I would bet the process "He's drawn a gun! OK, let me see what caliber it is. Hmm, it's a 22. OK, although I would run if it were a 38, I think I can take some 22 hits, so I'm not running." almost never factors into the thought process. I had a gun drawn on me once. I had just gotten into the passenger side of a car and closed the door when a punk on the street put a gun barrel right up to the window and pointed at my face. I didn't move, but the driver of the car drove off. Afterwards, I replayed it in my mind and suspected that the gun was actually either a pellet gun or maybe a 22 -- it certainly wasn't a 45. None of that calculation ran through my mind at the time.
significantly more individuals NEVER ceased to be a threat when engaged with .22 than .45 according to that article.
Number of people not incapacitated who took 22 hits was 31%. For the 45 ACP, it was 14% -- so a difference of 17 percentage points. But, again, the number of times a person needs to completely incapacitate an attacker in order to stop an attack is small. The average number of shots to stop an attack for the 22 was about the same as the 45 (1.38 vs. 2.08 -- probably within the noise of the statistic).
Also, total effectiveness of a handgun is not just probability of stopping an attack (let's call that Pstop). It is Pstop x (probability that you have the gun on you) -- call that Ptot = Pstop x Pcarry. Pstop for larger calibers probably isn't that much greater than a 22, but even if it were, Pcarry for larger calibers is probably much smaller than for 22's. As a result, when you are looking at a population and not just yourself (as you are willing to carry something larger than a 22 100% of the time), a 22 is a good recommendation.
And considering how small the chances of EVER needing a firearm are, if you're going to go so far as to carry at all, you ought to be prepared to go all the way in any eventuality.
All of your arguments apply to the recommendation to carry a shotgun instead of a 45. Of course, again, the Pcarry for a shotgun is much, much smaller than for a 45 or a 22, Ptot isn't very good, and so it isn't a good recommendation to the general population for carrying around for self defense. But, if you were willing to carry around a shotgun 100% of the time, then a shotgun would be a better choice for you than a 45.