TBI doesn't measure inlet air temp, it only measures engine temp. If the ENGINE temp is too low, the ECM esentially "chokes" the engine by adding more fuel than required, making it run richer than is actually necessary. I haven't looked at the PROM of a TBI engine in awhile, only thing I remember is that somewhere around 120* engine temp in some applications is where it looked like GM decided "choke" was no longer necessary by the way the PROM was setup.
Higher engine temps actually make more power, but you've got a bit of a catch 22 here...hotter incoming air keeps fuel atomized, theoretically leading to better efficiency in burn, while colder air is denser, theoretically leading to more power. So hot intake walls are a plus, but so is colder air. As much as any of us like to spout theory, the reality is without EXTENSIVE dyno testing of each possibility, which isn't really possible anyways, there is no way to know for certain. That's not even getting into the density vs. volume discussion keeping this post alive.
The G-tech mentioned is about as scientific as anyone has gotten so far it looks like. I tried flipped lid, single snorkel, dual snorkel, none of them did anything mileage-wise, which I check every tank. Different noise levels yes, more efficiency no.