Nick, you really have to LOOK at the images of Marangoni convection to get the argument. That's the IMPORTANT part of the argument. Moreover, the swirling patterns in the "swiss-cheese" formations DO indicate flow.
Ingersoll in the ABC News article gave the temperaure of the water ice *at the surface* as only a few dozens of degrees below zero. This is comparable to the surface temperatures in Antarctica. Quite important also is that some of the lakes in Antarctica with the shallow ice covers have high salt content:

APOD: 2003 January 28 - The Lost World of Lake Vida
http://antwrp.gsfc.nasa.gov/apod/ap030128.html

This alone can depress the freezing point dozens of degrees even without an ice over-pressure:

Making a Splash on Mars
"One thing we have to be careful of is our everyday experience that
water always freezes at zero degrees," noted Hoover. "It doesn't.
Water containing dissolved salts freezes at a significantly lower
temperature. Don Juan Pond in Antarctica is a good example. It's a
high salinity pond with liquid water at temperatures as low as -24
°C."
http://science.nasa.gov/headlines/y2000/ast29jun_1m.htm

In any case what's important is the similarity of the "swiss-cheese" formations to Marangoni convection formations. It really jumps out at you when you actually LOOK at it. It's pretty obvious that that's what's going on.

Bob Clark