In the course of China's rapid urbanization, the mismatch between public service provision and rural demographic shifts has become a critical challenge. This study investigates the systematic relationship between settlement size and service accessibility across 35,150 townships from 2014 to 2022. We utilize public facility coverage (PFC) to transition from simple facility counts to a more rigorous metric of spatial justice. The findings reveal a persistent superlinear scaling regime (beta > 1) in rural systems, in contrast with the sublinear efficiency typically observed in cities. Scaling exponents for healthcare (1.35 to 1.41) and shopping (1.42 to 1.49) indicate intensifying service polarization across township sizes. Furthermore, compared with eastern provinces, underdeveloped western provinces exhibit significantly greater concentration effects. These patterns are further associated with differences in local impact factors shaping variation in PFC across townships. Through a sixfold mismatch typology, we identify divergent explanations: administrative rigidity in healthcare leads to facility-led redundancy in shrinking areas, while market-driven shopping encounters service desertification. Comparison across mismatch types further shows that townships characterized by sustained or expanding facility provision tend to exhibit higher levels of PFC relative to population-based expectations, even under conditions of population decline. By identifying this persistent superlinear scaling, this study challenges the universal assumption of scale-efficient public service provision derived from urban systems, revealing a distinct logic of service polarization in rural hierarchies.