Generating power from very small temperature differences (2-3 degrees) with require gigantic surface areas to move energy. 25 mW/m2 is a TINY flux rate, and they are talking about how to get it to 0.5 W/m2? Need a bigger temperature difference...