The arXiv preprint (abs/2605.20232, May 2026) proposes a strikingly simple, model-free extremum-seeking law that lets wandering, black-browed, and grey-headed albatrosses extract energy from wind shear in real time using only local wind or energy-rate cues. Unlike prior optimal-control studies that solve the full trajectory offline, this mechanism operates as an online natural feedback loop, continuously adjusting lift to track the instantaneous extremum of energy gain. The authors validate it by matching both high-fidelity optimal-control solutions and published biologging records for the three species. Yet the work remains a preprint and relies on simulated wind-shear profiles rather than new field measurements; no sample size of individual birds is reported, only species-level comparison curves. This formulation resonates with earlier dynamic-soaring analyses by Richardson (2009, Progress in Oceanography) that mapped albatross GPS tracks to shear-layer exploitation, and with extremum-seeking control theory developed for autonomous vehicles (Krstić & Wang, 2000, Automatica). What the original coverage overlooks is the deeper cross-domain implication: the same low-dimensional feedback structure appears in bacterial chemotaxis and in gradient-ascent algorithms used in machine-learning optimizers, hinting at a convergent evolutionary solution to real-time resource maximization. Limitations include the assumption of steady shear and perfect local sensing; gusty or three-dimensional turbulence could degrade performance, a gap future drone tests could close.