The Trit master tensor X_{μν} [1] has three invariants: I₁ = Tr(X), I₂ = ½[I₁² − Tr(X²)], and I₃ = det(X). General Relativity is the I₁-only truncation of the full Trit field equation [2]. This paper derives the cosmological consequences of including I₂. The I₂ vacuum contribution — W_A^{vacuum}/Λ_c = ⟨Tr[A_{μν}A^{μν}]⟩_vacuum/Λ_c — is identified as the physical origin of the cosmological constant Λ_GR, which GR cannot derive. The T³ reflection symmetry of the Trit — the same symmetry that forces θ_QCD = 0 [5] — cancels the bulk QCD vacuum energy, leaving a residual that evolves with the cosmological B field. Because this residual is not constant, the Trit predicts dynamical dark energy with w(z) ≠ −1. Four structural predictions follow: w₀ > −1, wₐ < 0, dynamical dark energy preferred over Λ, and negligible dark energy at z ≳ 1. All four are confirmed by DESI BAO 2024 [3] at 2.6–3.9σ significance. The S₈ tension and the H₀ tension are identified as the same I₂ signal seen through different observational windows: one common cause produces three apparently separate cosmological anomalies. The Trit predicts these three tensions are correlated — a new falsifiable prediction directly testable by Euclid. The cosmological constant problem and the strong CP problem are the same problem, resolved by the same T³ symmetry. The absolute value of ρ_Λ and the redshift exponent n in w(z) are explicitly marked as open problems requiring the full 0Σ Cavalieri calculation [7].