Excellent work, and very revealing too.

However, I do have a little comment/question:

You set the temperature at 58F. This is a rather low initial temperature and does not take into account what the cabin temperature was at the time of the test and the mode the climate control system was working under. Correct?

Here's my point: By setting it at an initial temp of 58F there's a likelihood that you were driving the compressor at its highest load which imposes a measurable impact in to gas engine.
Now, assuming that you had the AC on Auto, the FE impact would likely vary dramatically because when the cabin temp reaches the low target of 58F, it would start to use the electric compressor to keep the temp, and thus impose a negligible load on the gas engine.

Because of this, some of us use the AC on Auto and never set the initial temp so low. Instead, we set the initial temp to a high value and then gradually we lower the setting so as to not impose a load on the gas engine. The gas engine will only power the AC compressor if the temp differential is too big. For small differentials the electric scroll compressor will do all the work, thus minimizing (not eliminating) the impact on FE.

Of course, taking this into account would really complicate things quite a bit, but your study remains a very clear indication of the impact of AC on FE.

Cheers;

MSantos