O2 could still get into the carboy by this method. If CO2 can get out by positive pressure inside the carboy, than O2 can get in when that pressure differential is reversed. It kind of a reversed bubbler then. If you were able to keep adding starsan to the airlock before the level of the outside column of starsan was sucked down below the holes in the middle piece of the airlock, then you might be able to avoid getting oxygen in the carboy.

I wonder just how much O2 gets pulled into a carboy during the chilling time. I know that quite a bit of liquid can be pulled up into a blowoff hose during that time, but I wonder if the amount of O2 in an equivalent volume of air is enough to really cause any problems. In one of the last session episodes, Chad talked about how he purges his hoses with CO2 prior to transfering beer due to the volume of air that he thinks can cause oxidation problems. If that bit of air would cause problems, then certainly the volume pulled into a carboy druing cooling could too, right? I realize that during transfering the beer is being agitated while the beer in the carboy isn't; and there's alos a blanket of CO2 over the beer in the carboy so maybe the comparison isn't valid.