Very interesting, I had not seen that study, in spite of a lot of looking. The manufacturer's claims for higher mileage are obviously not borne out by the tests. Since the engine does develop more HP with ethanol, I would have expected that for a given highway speed, the economy would have been higher than at the same speed using gas.
The only conclusion I can draw is that the injectors are putting substantially more ethanol in per power stroke, and that would indicate a very wide-band injector, which is entirely likely.
This makes sense, I guess, if the stoichiometric ratios for gas and ethanol are close to each other. If a shorter injector pulse makes the mixture too lean, then engine damage could result. Therefore, it is not possible with that arrangement to take advantage of the higher compression, since lots of fuel is being used to keep the engine from getting lean. Back to the drawing board, I guess. Thanks for the link to that study!