Resource theories have mushroomed in quantum information theory over the past decade. Resource theories are simple models for situations in which constraints limit the operations performable and the systems accessible. In a fixed-temperature environment, for instance, the first law of thermodynamics constrains operations to preserve energy, and thermal states can be prepared easily. Scores of resource-theory theorems have been proved. Can they inform science beyond quantum information theory? Can resource theories answer pre-existing questions about the real physical world? I will argue affirmatively, illustrating with photoisomers, or molecular switches. Photoisomers appear across nature and technologies, from our eyes to solar-fuel cells. How effectively can these switches switch? This question defies standard analyses, because photoisomers are small, quantum, and far from equilibrium. I will model a photoisomer within a thermodynamic resource theory. I will upper-bound the switching probability using thermomajorization, a resource-theory result that extends the second law of thermodynamics to small scales. Thermomajorization constrains the yield tightly if a laser barely excites the molecule, as in solar-fuel experiments. Electronic coherences relative to the energy eigenbasis cannot boost the yield, because modes of coherence have been proved resource-theoretically to transform independently. This work demonstrates that thermodynamic resource theories can illuminate nature, experiments, and materials. Reference: NYH and Limmer, arXiv:1811.06551 (2018).