Yeah, I love those government studies. When they actually test ALL models of ALL cars and publish the results, it might be usefull. So many variables. Only constants are the laws of physics.
Torque curve of a particular engine with a particular exhaust system is the first place to look. In my experience, most engines are most efficient near the peak of the torque curve. But Gas engines cvan be pretty good on HP vs. fuel consumption at WOT. Lot of the cars in the college mileage competitions will run wide open then shut off the engine and coast. Pumping losses of course go way down at WOT. Turbo motors are of course a different animal.
Now throw in CD, frontal area (nobody mentions that), friction losses in the engine and driveline. And for real fun, throw in the slippage between the tires and the road, especially at high speeds. Then add in the type of road surface; conncrete or blacktop. Blacktop has more slippage- part of what you are driving on is oil (asphalt is part of the remainder after you crack oil).
BC made some important points about the percentages each factor plays in the equation. Really you would want to do a spreadsheet with all the different factors involved and then chart it for vechile speed. Then to have some fun, move around the rear end gearing and you can tune somewhat for what speed you normally drive.
I had a Datsun 260Z back when dinosaurs roamed the earth (so says my kid). Thing way petty low CD and had small frontal area. Straight six, normally aspirated with four speed manual and and about 3.56 gearing. Would barely crack 20 mpg at 55 mph if I was lucky. At 85 mph, it could do 26 mpg. How? My guess is the pumping losses went down with the throttle opened wider and perhaps the airflow through the whole thing just hit a sweet spot. Drag through the air still went way up but how big of a percentage of the first number was it vs the second number? Notice in the chart shown above, cars generally get WORSE mpg at speeds below 50/55. That optimum point just moves around for different cars. Mercedes lists fuel economy for different cars in some of their data books. My '84 190D is expected to get 39.4 pmg at 90 kph but on 34 mpg at 120 kph. Don't know what it gets at top speed, but with only 72 hp, it won't crack 100 mph anyway.
I'd love to see the whole map of fuel consumption vs engine speed vs throttle position for each of my engines. What year was the SEL? I'll see if it is listd in my data book and see what MB says it will do. For what market (US, Japan, UK, DE) was it built?