Concrete data is concrete data. MetroMPG's rolling tests are going to be a heck of a lot more reliable than ones done at speed on the highway, where numerous extraneous factors like buffeting, wind direction and slope are harder to control. And lab tests are even more reliable, still.
-
Prizm versus Tracker: The Prizm is basically a Toyota Corolla. The Tracker is a Suzuki Vitara, which had much worse aerodynamics and a
completely different engine and drivetrain package with much higher driveline losses. Hell, rear-drive Suzukis with that motor barely put out 60 horses to the wheels. (Had one on the dyno for a turbo install before)
Suzuki's 1.6 SOHC is nowhere near being in the same league as the Toyota 4AFE, whatever application it was used in.
Don't see how that's a valid comparison at all.
Unlike the Fiestas, which you yourself provided, and which returned increasingly higher EPA numbers for the variants with taller and taller tires...
