Tru,
I've also been working on a not-yet-successful back end aero project (
http://ecomodder.com/forum/showthrea...upe-18906.html). I feel your pain in a good idea that didn't yet give good results.
If you have a gauge or a GPS that will read in tenths of a mile per hour, you could use this method...
I've been testing it with A-B testing. As aerohead (aka Phil Knox) suggested, build your mod as best you can, test it, then remove it and test again. I try for 2-3 good runs before I take it off, then another 2-3 good runs.
I've been doing coastdown tests. Find a stretch of road with a bit of a downgrade so your coast has some distance. Coasting between two consistent points, note your speed at the first and 2nd points. Start-point speed won't be identical every time but I think if you keep them within 1-2 mph of each other its good enough. Afterwards, calculate the mph speed lost for each run, and compare them. If all runs of each type (A + B) are pretty close in speed loss I'd consider it a good test. I've had cases where I threw out a run that was very different, because it was influenced by the traffic around me.
Check the "Details" weather forecast on weather.com to find a period when the wind speed and direction will be consistent through your test. And, do it when there are few cars on the road to eliminate their aero influence as much as possible.
This kind of testing isn't quick but I think it's faster and more informative than trying to just drive with it long term and see if you like the results. Adding a third group of runs so you have an A-B-A test is even better. In my case, each time I reviewed after running the A's and B's, I saw immediately that the mod hurt the coast, so I didn't continue testing. I went back to upgrade the mod instead. Still no "joy" but that hasn't kept me from doggedly pursuing this thing.