This paper presents the Dual-Field Interface Model (DFIM), a Scalar-Tensor-Vector Gravity (STVG) framework that conceptualizes spacetime not as abstract geometry, but as a physical, high-tension membrane separating two fundamental fields. Governed by a fully covariant Lagrangian action , the model preserves Lorentz Invariance via Spontaneous Symmetry Breaking in the vacuum state. It offers a unified mechanical resolution to three major open problems in modern physics: The Black Hole Information Paradox: By deriving a non-singular "stiffness horizon," the model maintains interface continuity. This allows entanglement entropy to naturally follow the Page Curve during evaporation, preserving unitarity without requiring firewalls. Dark Matter Candidate: The model calculates the mass stability limit for Primordial Black Holes (PBHs). It predicts that PBHs formed at the Electroweak Scale (~100 GeV) stabilize as saturated "frozen welds" with a mass of 10−5M⊙ (Earth mass), aligning with observational constraints for non-evaporating Dark Matter. The Hubble Tension: A Modified Friedmann Equation is derived from the field equations. This derivation proves that cosmic expansion is non-uniform: "stiff" high-density regions (galaxies) expand at standard GR rates, while "loose" low-density regions (voids) experience kinetically boosted expansion (Hvoid>Hcluster), resolving the discrepancy between Early and Late Universe measurements. Keywords: Quantum Gravity, Information Paradox, Dark Matter, Hubble Tension, Modified Gravity, STVG, Scalar-Tensor Theory, Primordial Black Holes, Analog Gravity.