Decision Fatigue is like muscle fatigue in that it is mostly due to the waning of glycogen feeding your brain. When the cause is lack of glycogen due to hunger, as in the case of judges who have good snack discipline, they exhibit it just before meal times. When the cause is lack of glycogen from high insulin due to sugary snacks (as provided by your typical startup) you exhibit decision fatigue after you eat candy or drink soda. Exercise helps because it it raises your resting metabolic rate so more glycogen is flowing through your body at any one time. Stimulants help because they increase hormones like epinephrine and norepinephrine which also increase metabolic rate.

There are secondary effects like nutrient deficiencies that can result in low essential neurotransmitters. And also if you stay focused on one thing, I think Troxler Fading can increase the the energy draw your brain needs to stay focused on that one thing.

Probably as important is complexity/uncertainty. If glycogen delivery is your capacity to work, complexity/uncertainty is the primary demand that the work has on your brain. So if you can keep redrawing/refactoring your problem in simple terms, I find you can keep going when others cannot.

It seems like you haven't read the articles that are linked to by the comment you're responding to.

Everything you said is precisely what is controversial and under question these days regarding both decision fatigue and ego depletion. The original experiments proposing such theories have failed to replicate.

Are you saying you can share experiments that show optimal levels of glucose in the body are not directly proportional to optimal decision endurance? I'd love to see something that refutes this, I can't imagine what it would look like when I have found it so easy to evidence the relationship in myself and others.