IIRC, back when the reformulated gas rules were being set up in California, the oil industry demonstrated a blend that could meet the rules cost-effectively without oxygenates like ethanol or MTBE. Ethanol makes it more difficult to meet the rules because it messes up the volatile hydrocarbon emissions, requires local blending at the distributor because it's unstable in the fuel and would separate between the refinery and the distributor. In addition, like methanol and MTBE, it makes the fuel leak right through common types of plastic and fiberglass tanks; more expensive materials and gas station tank replacements were needed to keep the stuff inside, for both water quality and VOC reasons (though most of that happened because of MTBE).

Only LA (South Coast Air Basin) is still, even, "maintenance" for CO - all the other former CO nonattainment areas in California have completed their 20-year "maintenance" period after becoming attainment and are now simply unclassified/attainment. IOW even in the worst case only LA still might need the stuff for controlling emissions from vehicles with carbureted engines - when was the last time you saw one of those in anything post-1990? In engines with a feedback fuel system (nearly universal now), oxygenated fuel (regardless of what is used; ethanol is nearly universal now) actually messes up operation due to making the fuel mixture effectively leaner; the system compensates by pushing more fuel through. Affecting, of course, fuel economy.

So yes, Bob is correct. The ethanol requirement is in fact not carbon-neutral when you consider the energy and chemical use of the farming and processing, or effective for emission control. It's basically pork for the corn industry in the Midwest.

On the train side, of course, it does produce some tank-car traffic, even unit trains or large blocks. The stuff doesn't ship well in pipelines either.