So after thinking about it last night, the biggest difference I can see between the picture above and Shadowcatchers set up is that SC's return is perpendicular to the evaporator coils. In the pic above I believe the return is parallel to the evap coils. I'm going to hypothesize that if the air enters perpendicular, then the velocity of the air would be strongest in the center of the coils, and slower around the sides, there might even dead spots in the corners. With the air flow being parallel then the air could be more evenly distributed. ( but of course I'm making up a story with the idea it might fit the evidence). I searched on line and I couldn't find anything that I thought helped understand what was going on.
One supporting theory I have is that when the air goes through the return duct, it would be at a slightly higher pressure, and as it gets closer to the evaporator those conditions would increase the likely hood that moisture could come out of the air. With the return aimed directly at the evap coil, then the moisture that comes out of the air would be directed directly at the cold coil. If its aimed parallel then any moisture that comes out of the air would be pushed towards the end of the plenum area by momentum,potentially reducing the moisture that went on the evap.
Again, this is all conjecture. If I had some high temps and humidity I'd build a couple of cardboard plenums and test the theory, but here in Socal, heat and humidity aren't due until late august and even then it won't be what some of you others will start seeing next month.
Tom
I did find this article, that implies that parallel airflow is better for reducing icing.
http://docs.lib.purdue.edu/cgi/viewcont ... text=iracc