I have removed the 15" wheel covers from my 2012 Prius since they interfered with accessing both the lug nuts and the tire stem (what a stupid, cheap wheel cover design), and replaced them with 2.5" center caps to protect the axle threads from corrosion. Although I didn't rigorously compare mpg before and after removing the wheel covers, on a recent out-and-back trip with the new center caps installed, the Prius averaged 56.5 mpg on a hilly, 240 mile course, at a speed of 66 mph (dry road, not very windy, air temp about 75F, cruise control employed, no hypermiling techniques). I was quite pleasantly surprised at the high mpg (even better than I typically get with my 07 Prius on that same trip), and can't imagine that the mpg could be much better with the wheel covers on.
It likely wouldn't be a detectable difference in a non-lab setting. There are hundreds of posts on the subject and no one has been able to prove anything one way or the other so take em off or leave em on as you see fit.
Math Nazi here: By itself, an average is meaningless. Only with a standard deviation / error bars does an average illuminate an answer to a question.
over at ecomodder, they claim something in the order of 4.6% reduction in drag at 60MPH with full moons. And I would have thought that partial covers such as they are for stock Prius were actually a good thing, considering brake cooling etc Experiment: smooth wheel discs tested A-B-A - 4.6% mpg improvement @ 65 mph - Fuel Economy, Hypermiling, EcoModding News and Forum - EcoModder.com and 65+ Vehicle modifications for better fuel economy - EcoModder.com
