Micro-scale waterfront spaces play a critical role in contemporary urban regeneration by supporting everyday activities and place-based experiences. However, existing studies often rely on linear evaluation approaches and insufficiently address the asymmetric effects of functional, environmental, and cultural attributes on residents’ landscape satisfaction. This study investigates the satisfaction structure of micro-scale waterfront spaces along the Grand Canal in Wuxi, China, with a particular focus on nonlinear demand mechanisms. A mixed-method framework integrating grounded theory, the Delphi method, and the Kano model was employed to identify key landscape attributes and classify their satisfaction effects. The results reveal a hierarchical satisfaction mechanism characterized by “basic–performance–attractive” attributes. Fundamental functional and environmental factors, such as accessibility, safety, water quality, and cultural authenticity, function as must-be attributes that primarily prevent dissatisfaction. Environmental comfort and social facilities act as one-dimensional attributes that linearly enhance satisfaction, while cultural narratives, memory-related elements, and ecological esthetics emerge as attractive attributes that significantly elevate emotional engagement when present. Sensitivity analysis further identifies priority intervention factors with the greatest impact on satisfaction improvement. These findings demonstrate the asymmetric nature of residents’ landscape satisfaction and provide a phased optimization framework for the sustainable regeneration of heritage-based micro-scale waterfront spaces, emphasizing basic reliability, experiential enhancement, and cultural resonance.

Sustainable heritage governance depends on institutional arrangements that sustain community participation and connect conservation decisions with local knowledge. This study examines how participation was represented, expanded, and institutionally configured in 86 Chinese national cultural heritage policies issued between 1982 and 2025. It integrates semantic network analysis, BERTopic modelling, and the Policy Modeling Consistency (PMC) index; the latter compares 11 purposively selected policies with different legal and administrative functions across nine design dimensions. The semantic network remained centred on cultural relics, protection, government, planning, and management, while communities occupied a peripheral position. Topic modelling showed that traditional-village, ecological, adaptive-use, and transmission concerns expanded after 2010 but were layered onto a persistent administrative-conservation core. Within the purposively selected 11-policy set, the mean PMC score was 5.18 out of 9. Policy instruments, evaluation, timeliness, and content were comparatively well specified, whereas X4 covered few of the predefined national-level institutional categories and X9 disclosure and feedback provisions were uneven. The 2024 Cultural Relics Protection Law scored 2.93 PMC points higher than its 2007 version (calculated from unrounded totals), reflecting broader explicit coverage of the coded provisions; X4 remained unchanged. These text-level findings support a bounded administrative-absorption proposition; the purposive PMC subset is not representative of all 86 policies, and policy texts cannot establish implementation outcomes, changes in formal decision rights, or actual community influence. Clearer responsibilities, closed feedback cycles, sustained capacity building, and procedural guarantees could strengthen the institutional basis for long-term community participation. The study distinguishes discursive expansion from institutional transformation and clarifies how policy design can support inclusive participation and the long-term continuity of heritage values.