I've seen a lot of these studies and the problem is always the same: the real-world spread on the inputs and outputs is always so much larger than what the report concludes. Since they're comparing ethanol and gasoline, you have two spreads to consider. As an example:
Worst-case for standard gasoline vs Best-case for corn ethanol:
A) Gasoline produced from Alberta tar sands syncrude (which has to be melted and partially hydrogenated using natural gas piped in from Northern Canada/Alaska) shipped by oil tanker (burning dirty bunker fuel) to California refineries for conversion to gasoline by distillation/cracking (with high CO2 emissions related to use of fuel to drive the require heating of the heavy crude oil);
B) Ethanol produced from corn grown with ammonia fertilizer made by splitting water with sun/wind energy, planted and harvested with renewable-powered electric equipment, and fermented and distilled into 95% pure ethanol (the eutectic limit of distillation) using only solar and wind power and in facilities right next to the farmer's fields;
And, conversely, the best case for gasoline and the worst case for ethanol:
C) Gasoline produced from regionally-produced light sweet crude with a minimal need for transport, hydrogenation or extensive processing, meaning much lower emissions from transportation and refining (we could even have electrolytic hydrogenation from water using sunlight to produce the hydrogen);
D) Corn produced in Kansas with large amounts of natural-gas-manufactured ammonia fertilizer, lots of natural-gas-derived synthetic pesticides and herbicides, planted and managed with diesel-powered equipment, then shipped overseas to China for fermentation and conversion to ethanol in a coal-fired ethanol distillery.
So if you break those spreads down you get a ridiculously wide range of estimates. Hence any specific number presented ('at least 25%' in this case) doesn't really tell you that much, other than if you dig through the specific methods used (which are generally hard to find, sometimes they'll put the methods in another paper hidden behind a paywall) you can tell what interest was backing this particular study. Some will even include things like the carbon emissions needed to feed the workers who grow the crops, or who work in the refinery, and so on, which is just nonsense. (The emissions of the restaurant workers who feed the oil refinery workers and the field hands have to be accounted for...?)
Personally, I think artificial photosynthesis is the vastly superior solution relative to either corn ethanol or crude oil gasoline, as you could plausibly run it anywhere with access to seawater and sunlight, saving agricultural land for food production (and conserving ever-scarcer potable water as well). It does however make sense to be able to convert excess agricultural production into useful products, although just composting it back into the soil might be wiser in the long run.