The paper studies a mathematical procedure called "finite group reduction" applied to certain complex systems of equations called integrable hierarchies. These hierarchies, named after Dubrovin-Zhang and Double Ramification, arise naturally in modern mathematical physics and geometry. They encode deep information about counting problems in algebraic geometry, specifically related to how curves can map onto geometric spaces. The main technical result is that when you apply a symmetry-reduction procedure (using a finite group of symmetries to simplify the system), the essential mathematical structures of the hierarchy, namely its two compatible geometric frameworks for describing dynamics (bihamiltonian structure) and a special organizing function called the tau function, are preserved intact.
The authors then apply this general result to a specific and concrete geometric problem: orbifold Gromov-Witten theory with a target space called the "pillow orbifold," which is a sphere with four special points each carrying a symmetry of order two. This is a well-studied object sitting at the intersection of string theory, algebraic geometry, and the theory of elliptic curves. The reduction procedure connects this theory to a simpler geometric space called the Hurwitz space, which parametrizes certain branched coverings of the torus.
A striking finding is that the Hurwitz space carries not just one but two distinct and mathematically interesting integrable hierarchies. This is surprising because the underlying Frobenius manifold structure, a rich geometric package encoding the algebraic rules of the theory, is the same for both. One of the two hierarchies turns out to be equivalent to something called genus-one topological recursion, a powerful and broadly applicable computational framework for extracting geometric invariants from a given input data. The coexistence of two inequivalent hierarchies on the same geometric foundation opens new questions about the relationship between different approaches to integrable systems in geometry.