Advancing physics unification within the Unified Forecast System (UFS) requires a process-oriented and scale-aware testing and evaluation (T&E) strategy. The DevelopmentalTestbed Center (DTC) supports this effort through systematic assessment of both operational and developmental physics suites, including the GFSv17 and GEFSv13 configurations.Since 2025, this work has expanded to evaluate the applicability of these physics suites within NOAA’s Seasonal Forecast System (SFSv1), which targets extended-range prediction at coarser spatial resolution. In collaboration with UFS physics leads and developers, this study focuses on key processes governing the persistent sea surface temperature (SST) warm bias in SFS prototypes, with emphasis on marine boundary layer (MBL) clouds and tropical shallow cumulus across scales. The analysis examines the roles of cloud microphysics, precipitation processes, and cloud-radiation interactions in regimes known to be particularly challenging, including marine stratocumulus, tropical active and suppressed convection. Sensitivities to model configuration-especially physics and microphysics timestep choices-are systematically quantified. A hierarchical T&E framework is applied across model configurations spanning horizontal resolutions from ~100 km to convection-permitting scales (~3 km), leveraging the Common Community Physics Package (CCPP) within a unified infrastructure that includes the Single Column Model (SCM), GFS, GEFS, and SFS applications. Results demonstrate strong regime-dependent sensitivity of cloud–radiative feedback to both physics parameterizations and timestep configurations, as well as persistent deficiencies in representing MBL cloud regimes at extended lead times. These findings provide actionable guidance for improving process fidelity and advancing unified physics development across UFS applications.