I suspect the vehicle (and driver, and environment) of interest would need to be tested. There are probably way too many variables to make a useful blanket statement.
Before the Versa I had a Cabrio; its 2L was the only VW engine rated for regular gas. Back in 2007-2008 I ran a test on Shell Regular, Plus, and "V-Power." It was real-world, not tightly controlled, and not blinded, so it hardly qualified as an "experiment." But I ran the test for a LONG time to try to get reasonable data.
In my specific car (with me driving, etc.) I saw a slight FE improvement with premium, but not enough to offset the additional cost.
The data is available if you're curious, but it's on another forum and I don't know if I'm allowed to mention it here. Hardly worth looking at anyway; the improvement in my case was very subtle.
Rick
|