Accurate real-time range prediction in Fuel Cell Hybrid Electric Vehicles (FCHEVs) is frequently compromised by simplistic models that neglect the stochastic nature of real-world driving and the critical coupling between component degradation and energy consumption. This paper proposes a novel, unified, multi-physics, health-aware framework that integrates a deterministic physics-based powertrain model with empirical degradation model for the fuel cell and battery. Central to the architecture is a multi-objective power delivery system (PDS) that utilizes a descriptor-driven fuzzy inference structure, this research proposes a Unified Fuzzy Grey Wolf Optimizer (UFGWO) based strategy. Unlike isolated component models, the proposed approach adaptively captures the immediate performance impact on the health of the fuel cell and battery. The results demonstrate that the UFGWO strategy presents an economic and cost-effective strategy for the robust performance of FCHEVs across varying driving conditions. Furthermore, the framework incorporates a strategic Environmental, Social, and Governance (ESG) assessment, leveraging the traceability of physics-based modelling to meet increasing regulatory transparency requirements. Experimental results across diverse drive cycles from urban utility to heavy haulage, demonstrate the hierarchical update structure of proposed PDS provides a superior exploration-exploitation balance, effectively maintaining the fuel cell in high-efficiency regions while mitigating battery stress. The proposed strategy achieved average reductions 82.48%, 56.03% compared to traditional Rule-Based and Charge-Sustaining benchmarks respectively. Moreover, the UFGWO outperformed Particle Swarm Optimization (PSO) with overall improvement of 33.70% – 55.20%, establishing it as a robust, ESG-aligned solution for next-generation sustainable mobility.