Effect of non-ethanol fuel on indicated vs calculated mpg ?

When refueling, I usually have a displayed mpg of 56-57 and a calculated mpg of 53-54, so typically a 3-4 mpg difference. With non-ethanol gasoline this difference seems to dramatically decrease. I first noticed this effect when driving in Alaska where non-ethanol gasoline is the norm two years ago. But earlier this Summer I refueled with non-ethanol fuel at a Stewart's Sweet Shop in upstate New York, and 550 miles later when I refueled the MFD displayed mpg was 60.1 and the calculated mpg was 59.9, a surprisingly small discrepancy. That brought back the memories of a similar shift using the non-ethanol fuel in Alaska, where I ultimately had to recalibrate the fuel consumption estimates of my Scangauge II in order to have the estimates better conform to actual consumption .
Has anyone else noticed a decrease in the discrepancy of displayed vs calculated mpg in a Gen III Prius when using non-ethanol vs 10% ethanol fuel? Thanks in advance for relevant info!