boy... you are going to sprain something reaching so hard.
Not reaching at all. Just trying to illustrate that the low gas light and tank full nozzle switch are not accurate measurement devices. My main point, though, is that the fuels being compared are not sufficient to isolate the effect of ethanol on mileage.
a) My car said "full". My gallons in were roughly the same (about 12.25 gallons) as every other time they low gasoline light came in.
I have never had refills that consistent. I guess I don't always refill exactly when the light comes on. Of course, I don't drive often these days, and haven't for 10 years. I find that level of consistency, refilling to +/- 0.05 gallons quite surprising. Anyway, you never specified that you recorded the fill amount at the E0 pump, only the next refill after. Actually, you never specified whether you took the measurement before or after using a given fuel. Unless you did both and calculated the exact mileage for both adjacent tanks (rather than just some random E10 tank), it's hard to say. The error in "around" 12.25 and "around" 265 or 300 miles can be significant. 261 miles / 12.4 gallons is 21 mpg, 270 miles / 12.1 gallons is 22.5 mpg, and 295 miles / 12.5 gallons is 23.6 mpg. It is even more so if around 265 miles means between 250 and 275, and around 12.25 gallons means between 12 and 12.5.
Not 35 miles difference. You get to the point in mileage where your mileage light "always" comes on and you notice it's not coming on...
IT'S NOT A 35 MILE DIFFERENCE. There is less energy density in ethanol than gas. It's like a 20 mile difference. That's about 7%. Your methods are not nearly good enough to measure a 7% change. Not even close. Especially in 3 trials.
Or using your expectation of up to a 10% reduction, it's a 5 mile difference, which is hardly significant. 10% less than 300 miles is 270 miles.
Yes, my expectation from Keith Knoll's report is that I'd lose 3.5% mileage per tank using 10% ethanol. Reality is not matching that expectation. Hell, my worst case expectation would be that I would lose 10% with ethanol but my observed loss is over 10%. Having ethanol in the tank provides worse gas mileage for my honda element than the mystery substance I'm going to presume was 100% gasoline (since super and premium do not raise my mileage).
100% gasoline doesn't really mean something specific. It just means it's some mix of liquid hydrocarbons with some limits of how much of certain ones are in it. It is only different from other "non-100%" gasoline in that it presumably wouldn't contain additives beyond what comes out of the refining process. However, they might just mean ethanol-free, and who knows what it would contain. Or maybe the E10 has other components such that it is not 90% regular gasoline. Maybe it has less alkenes because ethanol makes them evaporate; alkenes have higher energy content than the rest of the fuel. Who knows. Since they don't advertise the chemical makeup of the fuels, it is impossible to know unless you mix and chemically test the fuels yourself.
You can only expect the 3.5% reduction if the E10 mix is actually 90% the-same-gasoline-as-E0 and 10% ethanol. All we know is that it is at most 10% ethanol, and 90% other stuff. Hell, it could be 85% gasoline, and 5% stuff to keep the ethanol mixed. And that 3.5% is specific to the fuels they were using; the actual amount depends on the energy density of the particular batch of gasoline, and could be higher or lower.
You should not expect super or premium to raise your mileage, unless your car requires them. It generally just means higher octane, which means less energy than regular. They generally contain more additives (like ethanol or MTBE) to raise the octane rating, and those additives are probably not actually octane.