I believe you're taking into account conversion losses but forgetting to fully take into account ICE losses and here it is why: Imagine 2 TCH driving for an hour at steady 40 miles per hour on flat road. One car uses ICE at all times, another switches between ICE and EV. For simplicity lets also imagine the ICE needs some minimum amount of energy to run, which is totally wasted and the remaining energy is 100% efficiently used to either charge the battery, push the car or both. Now the car with ICE running all the time will waste more energy (minimum energy needed to keep ICE spinning x60 min) than the car that had ICE running part of the time, let's say minimum energy X 40 min. It is impossible for me to say if the conversion losses are higher than ICE friction losses from running 20 minuttes longer without some elaborate tests. If I had to make wild guess, I would think less friction loss from ICE not running all the time would just about cover the conversion loss with some particular cases going one way or another. So IMO it doesn't matter if you keep forcing EV mode or not, at the end it won't make much difference, so maybe why bother? I do agree that changing in driver behavior has much bigger effect on FE than the type of mode the car is in by itself.
Tell me more about the scanner, Ive been looking for something like that. Does it have specific TCH functions or just generic OBDII?