There is a 2004 paper by Lynd & Wang of Dartmoth published in the Journal of Industrial Ecology (an MIT and Yale publication)showing EROEI for corn and cellulosic biomass in different types of process configurations.  To tie everybody in the paper was guest edited by an Iowa State University Engineering person.  They were very thorough in identifying the energy inputs from fertilizing the crop all the way through crop handling, enzyme treatment and cogeneration in the ethanol plants.  They show positive EROEI for most processes.  Yes, corn can result in negative EROEI if set up indcorrectly and yes cellulosic can be better in some configurations.  But corn and cellulosic are about equal if configuring ideally for the respective feedstock.  The key is you can't set up one type of operation and then change feedstock or process parameters. No one size fits all.  Physical plant construction is critical to positive EROEI.  Sounds a lot like refineries and sour crude issues we discuss here all the time.

There have been enormous strides made in industrial enzymes for ethanol production.  This has been led by Novo Nordisk but others are working hard in that area. The enzymes are required to lower the energy cost to breakdown the complex molecules into sugars that yeast can convert to ethanol.  As I posted earlier today there are also growing markets for the edible waste from the ethanol processors, particularly distillers grain from corn.  Not only have the EROEI been shown to be positive there are very strong economic incentives to cycle corn through ethanol plants before feeding animals.

The point of all this is that old data in the biofuels area is not to be trusted.  People are figuring out how to get positive EROEI in a sustainable way.  I sometimes wonder if the scrutiny applied to biofuels was used for petroleum from exploration to wells through refining to cleanup and waste disposal what the true EROEI would be.  Those considerations are typical for evaluating ethanol.