In low-altitude public aviation, aerial emergency-response SoS design must determine fleet composition and basing under multi-scenario constraints, with decisions that remain explainable and traceable. However, mission-level simulation and SoS fleet-and-basing optimization are tightly coupled: changes in environment, fleet composition, or basing propagate across levels, resulting in expensive nested multi-objective search, limited model reuse, and weak end-to-end traceability [1,2]. We propose a layered decoupling framework that embeds Minimal Operational Units (MOUs) within an RFLP structure [3]. R defines effectiveness and cost evaluation metrics and constraints; F maps mission workflows to executable evaluation models; L standardizes MOUs as parameterized capability packages with defined interfaces; P instantiates a discrete Pareto-optimal mission-level MOU library [4] for SoS-level layout optimization. Layouts are explicit compositions over the library of evaluated MOUs, enabling cross-layer traceability. A forest firefighting case study shows that Pareto-efficient MOUs can be grouped into several patterns: formations of 4–5 10-tonne-class helicopters deliver the best effectiveness–cost trade-off. In a nationwide application for China with 363 added helicopters, coverage across priority fire-risk zones (Levels 1–3) improves by up to 39% over the 2025 baseline. Overall, the MOU bridge converts nested mission–SoS optimization into scalable combinatorial search over a finite discrete library, enabling computable, traceable, and engineering-feasible concept generation.