Yes, the efficiency gauge is essentially what I have been trying to match up with expected mi/kWh directly.
I think I have figured out that the root of the variance issue however is thermal management of the battery. I just did a run in 61F vs. the 78F temperatures earlier and there was a significant difference in fuel economy, netting over 40 MPG this last time around over 79 miles at the same 65 MPH average. That is more inline with expectations for it on gas only. (And it's possible I got a short fill before my first test run as well.) Unfortunately the only way I have to control for battery thermal management is ambient air temperature during testing. Fortunately for at least 6 months of the year around here, 61F usually happens once or twice a day, but depending on the right weather still sucks and I really just need a long fast hill to coast down!
Anyway, on another note, it looks like the Torque app has had something potentially useful added to it with fuel consumption reporting. The value reported is completely wrong, but it does accumulate a long term average which settled at 17.8 for the recent test. Given the 40.8 MPG actual measured fill, I can probably safely
change the displacement in the settings for a more accurate idea of the extended consumption for each trip as quick way to compare without stopping to fuel up.
I suspect it will be less vulnerable to inaccurate fuel consumption since the generator runs a fairly constant load.
Oh well, making progress even if I have yet to really test any mods!