The path-following and formation control of unmanned surface vehicles (USVs) in complex marine environments has attracted increasing research attention in recent years. Despite advances in line-of-sight (LOS) guidance, artificial potential fields, and disturbance compensation, existing approaches still suffer from issues such as unstable convergence under fixed lookahead distances, abrupt repulsive forces and local minima, and oversimplified modeling of wind–wave disturbances. To overcome these challenges, this study proposes a dual-layer cooperative control strategy that integrates a lateral-error-driven adaptive LOS guidance law, an improved artificial potential field, and wind–wave disturbance compensation. At the kinematic level, the adaptive LOS improves path smoothness, while the dynamic control layer combines disturbance compensation with the improved potential field to ensure robustness. Lyapunov-based analysis demonstrates that the closed-loop system achieves input-to-state stability (ISS) under compound disturbances. Simulation results show that, with a lateral error convergence criterion of ±0.1 m, the proposed method guarantees effective convergence of all USVs, enhances formation stability, and significantly outperforms conventional LOS guidance, thus verifying its effectiveness and engineering applicability.