I didn't totally get your last phrase on the comparison (I think you mean we evolved less rational low power state on low food to conserve energy for when we need it more, because low power but operational is more survivable than passing out due to hunger - did I get you right?), but this is a good counter. How could we test between them? It's possible it's /both, but without a good experiment hard to say. Hypothesis 1: brain on low sugar, increases variance in problem solving. Hypothesis 2: brain on low sugar low power mode prioritizes simpler solutions (I think I got you?). Your point is a good counter to mine, but I think the qualifier is that in low food situation, a less rational state is a more effective search algorithm to find key resources, than a more rational one (with definition of rational tempered by what is achievable passing as rational for most humans in aggregate). But maybe we are getting too complex for a hn comment when we might not be running the experiemnt oursleves hahaha :)