It's not confounding variable because this "variable" is actually constant in the entire group; the comparison is not fasting muslims vs non-fasting non-religious judges, but muslim judges with longer vs shorter fast or fasting at different hours
It ads a lot more variables making it unsuitable, day time is not strongly associated with hunger. For instance its possible that the 5 prayers during the day have effect. And also the feeling of hunger is not constant, it might be stronger around normal eating times etc. There are just too many variables to conclude anything.
