The other issue is that all of the 'studies' are done under laboratory conditions, nor 'real world' conditions.
As I mentioned above, Ethanol is very hygroscopic, when mixed with gasoline it allows the gasoline to also (albeit to a lesser degree) allows the gasoline itself to become hygroscopic.
I've seen Goobermint test results of in ground tank samples showing an large quantities of WATER in solution in the gasoline.
When you add the losses, financially and energy-wise, of alcohol PLUS water, the end result can be a rubber-to-road power loss of 20% to 25% or more.