Are you sure that number is current? According to https://www.wired.com/2011/06/five-ethanol-myths-busted-2/, ethanol production has an energy balance of 8.8 MJ/L (Megajoules/liter) and (though they leave this number out of the article), ethanol has about 20 MJ/L (I don’t know why there is a range given in https://en.wikipedia.org/wiki/Energy_content_of_biofuel - is that from water contamination or other hydrocarbons being in the mix of ethanol bio fuel as opposed to lab grade ethanol?). Gasoline is around 47 MJ/L so if it took 92% of a gallon of gas to make a gallon of ethanol, you would definitely not have a positive energy balance (and I agree it is a stupid thing to do at scale at that point, unless you can argue for the reduced pollution at high altitude cities from the oxygenation effect of using ethanol as an additive is worth the extra gasoline consumed.
There was that 60 minutes piece a few months ago that looked promising for improving the efficiency of cellulosic ethanol and I believe the inventor (who had a scheme I never would have thought of - using electron guns to perforate the cell walls of the cellulose to make the biological process used to convert the material to ethanol work better) claimed it could with unused agricultural material supply up to 15 or 20% (I forget the number - it wasn’t half that’s for sure) of current oil consumption. Could be hype, but we’ll see. But even now it doesn’t seem like making ethanol from the current process is such a bad idea (most of it gong to oxygenated fuel I believe and it does help with smog).
@Trog, any thoughts on this?