A good experiment would be start the circuit at 85% full battery (or 100% - 2.4 kWh) and keeping the return up the hill at the end of the circuit. It would cool to know if the lost 8 miles show up.
This experiment seems a little tricky to me. Just to illustrate, let's make some numerical assumptions.
First , assume that the inherent EV range is 48, but the hilltop scenario is reducing it to ~40. The hypothesis is that Dan's
GOM is telling him 40 because it learns this hilltop scenario.
Now, instead of charging to 100%, you start charging to 85% (in order to have room for the downhill regen). The GOM should indicate .85*40=34 miles (It hasn't learned anything new yet). This would need to be repeated for several consecutive days, but if every day, you start with 85%, then the GOM will hopefully learn that there are ~8 miles gained because of the previously lost regen that is now being captured. The starting EV range should increase by ~8 miles (increasing from 34 to 42) with the 85% charge. This would mean that a "full" charge 'could' yield ~42/.85=49 miles.
After the experiment is done, then charging to 100% should show this ~48 mile range at first (until the GOM once again re-learns the old scenario and knocks the estimate back down to 40)...
I think this is a very good experiment, but it is tricky to perform it in a way that yields convincing results.
If this proves out, then it would behoove Dan to always start with 85% charge so he can capitalize on the initial regen that he's loosing right now...