how are you calculating your energy expenditure... you're not just going off the "guesses" that MFP provides for cycling exercises are you? because, they're notoriously rubbish.

Get a HR monitor, and a Garmin, record the rides, use the Calorie Calculation the Garmin gives. Okay, so you're "off grid" in the bush - there's plenty of sunlight, so presumably you've a solar charge panel to keep them topped up, along with the phone...

that's going to be so much closer than any guesses that MFP comes up with - because MFP doesn't know how heavy your bike is, how windy the ride was, what tyres you're rolling on, and what surface you're rolling them on. That's 4 variables that can all double (or more) the wattage needed to move at a given speed. Hence, I keep referring to MPF figures as guesses.

I was really interested in your question/observation here, because for years people have complained that the MFP exercise calorie estimates are too high. I wonder if the algorithm has changed.

I personally don't use MFP's calculations, instead taking an estimate from my Garmin which knows my height, weight, age and crucially, my heart rate during the exercise.

This last point makes for a much more accurate estimation, which of course, is all any of these "calculations" are - estimates based on known baselines to which your individual information is applied

Forgive me if I am explaining things you already know.

I can't find any information about the algorithm MFP uses to estimate caloric burn. The fitness industry has for decades elected to over-estimate such things, presumably to give the user of that particular exercise bike or treadmill a greater sense of satisfaction. In more recent times, as we've become more health conscious, we've been demanding more accuracy.

To your specific problem here, as someone else suggested, an activity tracker that is truly measuring your effort is going to get a lot closer than a less informed calculation from MFP.