With the caveat that I know far less than Dale Atkins does about this subject...
It seems like the argument in the article is that "swimming" is bad because you aren't using your arms to create an airspace, not because "swimming" is bad per se. The flaw in the argument, it seems to me, is that "swimming" is more likely to keep you on the surface and therefore make you not need an air space. The article says:
By "swimming" you make yourself more like the big corn flake that's up at the top rather than the crumb that's at the bottom - or to go with the nut analogy, more like the Brazil nut and less like the hazel nut. Here's one experiment I'd like to see: Get 50 crash test dummies with their arms and legs spread out in an X and 50 crash test dummies in fetal position. Put them on top of a slope that's about to go, trigger it and then see how many of each kind stay on top. I'll bet that there's a marked excess of the dummies with their arms and legs stretched out in an X.
This is ignoring the effect of actually moving your arms, but I'll bet that simply the fact that "swimming" makes you, in effect, a bigger particle makes "swimming" beneficial. And it's probably a whole lot more intuitive than popping out into a stiff X shape when you're caught in an avalanche.