Question I was considering was whether evolution had a goal in this? I guess it did. Nervous-system-noise helps explore that state space and extreme states (hunger, sleep deprivation) lower inhibition, leading to greater exploration of the decision space. Biochemical/genetic mediators of more efficient (in aggregate across all humans ever) search algorithms.
Those despotic genes! Programming us, the nerve of them! :) Hahaha :)
In terms of whether that's large enough for a ruling, I guess it'd depend on all the pertinent factors: the rule, the judge, the context, the level and nature of discomfort, and the definition of significant. Within that multid space, I think, contrary to you, there's ample chance to produce large effects.
Mitigating this is perhaps how similar many judges routines are. What might really be strong support would be some control, and putting some jurists on more extreme diets/routines hahaah! :) Of course, only in mock trials tho, as otherwise that would be unethical.