Fuel economy vs. grade and brand

There have been several threads recently that suggest considerable complexity in determining whether fuel choices are impacting fuel economy. I don't know what the answers are, but it appears to me that we could benefit from a reasoned discussion/approach to tackling this one.

Sampling the various threads in different sections of the forum: Some folks have gotten very poor mileage with 10% ethanol, some not so bad. Some folks have reported much better mileage with mid-grade or premium, some have gotten worse. Some have gotten great mileage with branded refiner's gas such as Shell.

Trying to sort this out and make sense of it I'm proposing the following factors to consider:
1. 10% ethanol. This has a lower energy value and should result in reduced mileage, but it is slightly cheaper per gallon.
2. Toyota designed the car for 87 octane unleaded and if everything is in spec and matches their design temp/pressure range this would theoretically give best efficiency and lowest cost.
3. Environmental factors may favor one grade or brand over another (sea level in summer vs. high altitude for example.)
4. There can always be variability in the manufacture or set up/wear of the components whether or not they are all still in spec--valve timing could be shifted slightly/tolerances tight or loose, ignition timing could be advanced/retarded slightly, O2 sensor or MAF might be slightly biasing the mix, the plugs might be a little hot, the injectors (or one injector) might be a little off.
5. Certain refiner branded gas might very well have specific components (particularly for octane or regulatory targets) that outperform generic/lowest cost materials (see Murphy's at Sam's/Wal Mart?)
6. Obviously cost is a factor, so going the cheapest per tank route is not always best, nor does running premium necessarily make sense even on vehicle where it might give slightly higher mileage.
7. Driving habits might play a factor. For example, I try not to let the ICE lug as I believe this to be inefficient, so I'll try to get it to kick out, or apply more throttle to move the load toa point that doesn't feel like it is lugging.
8. The different Prius models will probably respond differently.

Essentially, the only way to determine whether or not your gas fills are contributing to trouble or are at optimum is to experiment somewhat. On the surface it would be tempting to proclaim that Toyota says 87, and we know that 10% ethanol has some negatives, so choose 87 with no ethanol. But that might not apply if your valves are a little tight or ignition timing a little off, etc. You might find your vehicle runs best on mid-grade or even premium.

So maybe we need to expand suggestions in fuel mileage discussions in late stages to include running test tanks of Shell (or other respected brand) and/or higher octane to see if it makes a noticeable diffrence. Doing so could indicate a problem with a local station/brand or that there is something about the car that should be examined. This latter part could be particularly helpful, especially if one detects signs of ping/knock.

Feedback and discussion are welcome as this is a compilation of ideas. I would like to turn it into a sort of troubleshooting addendum.