Regarding my future tires, I have made a graph showing FE as a function of different tire sizes. They range from 145 to 185 in width, 55 to 80 in height ratio and 13" to 14" in rim diameter. These are all the sizes I could mount (and that are at least as big as my current tires), not all the ones I consider to mount (I’ll try to avoid 175 and 185 for example).
y : FE in l/100 km
x : tire sizes arranged by diameter, the letter at the end in the rolling resistance class : A is best G is worst. My OEM tires are E, they are all at least as good except one (F). My current tires are on the far-right.
The FE shown is the theoretical FE at 56 mph with as only mod longer 5th gear (estimated gain of 10%). It takes into account the changes in rolling resistance, Cd (width dependent), frontal area (width and diameter dependent) and RPM.
For the Cd, I took the most optimistic figure I found in literature : +0,002 Cd for +10 mm tire width.
For RPM impact, I estimated a 7,5% FE gain for a 10% increase in diameter. While this is probably near to reality for tires that are 0-10% bigger, I imagine that for 20% bigger tires, the gain will be inferior to 15%. So FE for the biggest tires are to take with caution. I won’t be able to mount such tires anyway, I have yet to test with an old 165/70/14 tire what is the maximum I can mount. And then, I’ll make my choice considering the price too.
Also, this is only for changing the front tires. Since I have much more room for bigger tires on the front than on the rear, I’ll keep my original tires on the rear and put bigger tires on the front for lower RPM.
The graph shows that diameter seems to be the main factor in FE change. The ones that are out of the trend are tires that have either particularly bad RR or are very width.