"Across all experiments, individuals with high self-control were seen as more powerful, and as better suited for powerful roles, than individuals with low self-control."

Part of me sees this research as "doing everything by the book" and consequently failing to capture important insights. For instance, every economic model of humans assumes that we're "rational agents", yet we know that people with poor impulse control (like ElMo, Trump, etc.) are perceived as powerful by a wide swath of the public.

Maybe their "turns down dessert" thought experiments are just that -- thought experiments -- and do not reflect what humans actually do, IRL.