I do not mean to suggest that regen is 80% efficient, or that the addition or removal of bars corresponds exactly to SOC. These topics have been covered on other threads.
I was simply suggesting a way to correct the bars/range calculation estimate for altitude. Obviously, it will be inexact, due to road variations and "micro" regen/acceleration events due to traffic, curves, dips and rises, etc., and likely more accurate over longer drives.
for example, my "commute" is a 50 mile round trip, beginning with about a 2000 ft (1500 net) descent in the first 7 miles. If I start with 10 bars or less, the first bar typically disappears at about 14 miles, then 2 more are used in the next 11 miles (flat at 55-60 mph).
The return trip uses the same 2 bars for the first 11 miles, then 3 more for the remaining 14 miles.
On longer trips, the estimate has also worked for me.
For example, on my first drive to Shasta NP, it took 10 bars to cover 42 miles and about 5,500 ft (5,000 net) of ascent, but only 1 bar was consumed on the same route while descending.
Remember, I'm not suggesting regen is 80% efficient. When you are descending on a (not too steep) road at 45-60 mph, most of the time you are applying light "throttle", and are not in regen mode, which you should avoid for best efficiency and range.
While regen is probably a lot less than 80% efficient, my experience is that you can typically regain about 80% of the total energy consumed in your ascent, when you descend on the identical route.
Some of you with SOC meters could probably expand on this.