Not to mention, the surge of glucose from eating the sugar cubes could've very well affected sentiment and responses. Daniel Kahneman talks about this in his book, specifically how groups given a chocolate bar in between tasks performed better than control groups, presumably due to the brain's dependence on glucose, and its fatigue in the lack thereof.
Recent evidence suggests that glucose depletion doesn't seem to affect the brain. For example, [1]. Moreover, the calorie consumption of the brain is basically constant regardless of what it is doing. I would hazard an untrained guess based off of these that the difference is in changes due to the body knowing of consumption of glucose versus any direct effect of the glucose itself. See also the study that athletes performed better after swishing a sweet drink in their mouth and spitting it out just as they did when actually drinking it.
Yes, thanks for the correction! And for the sake of posterity, it's "Does the Brain Consume Additional Glucose During Self-Control Tasks?" by Robert Kurzban
