The nature of dark matter (DM) and dark energy (DE) is traditionally addressed through the introduction of new fundamental fields or modifications of gravitational dynamics. In this work, we develop an alternative framework in which dark-sector phenomena arise as emergent, operational effects of access limitations and entropic structure within modular quantum dynamics. The central premise of the approach is that physical description is intrinsically relative to an algebra of accessible observables AF . Degrees of freedom outside this algebra remain operationally invisible, yet can influence correlation structure and emergent geometry. This leads to the notion of access-relative physics, in which hidden sectors manifest not through direct observables but through indirect geometric and dynamical effects. Within this framework, we introduce a canonical MaxEnt reference state σF (ρ) that defines an entropic vacuum consistent with accessible information. Deviations from this vacuum are quantified by the relative entropy functional VF (ρ) = D(ρ∥σF (ρ)), which we interpret as an effective excitation energy relative to the access-defined vacuum. This construction enables the definition of an operational vacuum parameter Λeff , capturing vacuum-like contributions in a coarse-grained description. We show that curvature, entropy dynamics, vacuum structure, and hidden-sector effects are linked through a unified geometric–entropic relation. Within this relation: • entropy acceleration generates curvature, • entropic vacuum produces a background geometric contribution, • hidden correlations induce residual backreaction. This allows for a natural operational separation between DM-like and DE-like regimes without introducing new ontological entities. Dark matter corresponds to access-invisible backreaction, while dark energy corresponds to stable entropic vacuum drift. The framework is supplemented with a reproducibility-first protocol, explicit diagnostics, quantitative stability criteria, and failure-domain reporting, transforming it from a conceptual proposal into a structured and testable theoretical program.