economy better in the hills?

[steeve]

Please tell me if I am wrong. First, I am a mech. engineer by vocation and have always believed I get better fuel economy in hills or low elevation mountain driving. to prove my point with a disbelieving friend, I recently did an experiment; I drove my 08 3/4 ton Dodge/Cummins diesel truck down the mountain from where I live, starting at the top at a marked point, where I re-set the OBD-2 fuel mileage display, and checked the average mpg at the bottom of the mountain; it read 59.2 mpg for the mostly downhill, 10 mile long section,(very little, if any, throttle use was necessary). on the return run home, I re set the display at the same ending point of the first run and drove up the hill to the original beginning point, so as to duplicate the same distance and elevation changes as the down-hill run. the display read 17.4 MPG for the up-hill leg. I averaged the two runs out at 38.3 mpg. My usual average for flat land driving is between 24 and 25 MPG with no wind influences. So, it looks to me that my belief that I get much better economy in the hills is correct. After trying to explain this to my friend: (since up-hill mileage is only slightly worse than flat land economy, while down hill economy is vastly better than flat land economy); when you combine the two, the average of the up-hill and down-hill is better than driving on the level. I should add that my speed during the down hill leg was governed by what I could safely/comfortably manage around the twisty road corners, while on the up-hill leg, I did not add throttle in the steeper sections, thereby causing the truck to slow down a bit more than most traffic would travel, but not so slow as to inconvenience any following traffic. The net effect was that I went down the hill 10-25 mph faster on average than up the hill. I also pick a "racing line" through the corners,(otherwise known as "using the whole road"), which minimizes the need to use the brakes allowing me to carry more speed through the corners which requires less or no throttle coming out of the corners.
My friend, who was a long-haul truck driver, would have none of this argument, flatly stating that my experiment was flawed in some unknown way. But over this 20 mile test distance of completely equal amounts of up and down hill driving, it appears that I got 38 mpg, which is double what I can get on flat roads. I would very much appreciate comments from you veteran hyper-milers. Am I right or am I missing something?