Closing the Cosmological Lithium-7 Problem through a Recognition-Cost Suppression of 3He(α,γ)7Be

The cosmological lithium-7 problem is a factor-of-three excess of the standard Big Bang Nucleosynthesis (BBN) prediction Y^BBN_Li ≈ 4.7 × 10⁻¹⁰ over the Spite-plateau observation Y^Sp_Li ≈ 1.6 × 10⁻¹⁰, persistent across eight independent surveys spanning four decades. We derive a closed-form suppression factor for the dominant lithium-producing channel ³He(α,γ)⁷Be from a single dimensionless cost function J(x) = (x + x⁻¹)/2 − 1 on the Coulomb-barrier ratio. The compound factor η_RS = (1 + J(φ))(1 + 1/45)φ² takes the value η_RS ∈ (2.93, 3.04) from analytic bounds on the golden ratio φ = (1 + √5)/2, with central value 2.973. The prediction Y^RS_Li = Y^BBN_Li/η_RS = (1.55 ± 0.04) × 10⁻¹⁰ sits inside the cross-survey median observation Y^Sp_Li = (1.59 ± 0.06) × 10⁻¹⁰ at the 0.7σ level, with no free parameters. The discrepancy is closed without invoking stellar depletion, varying constants, decaying particles, or modified BBN reaction rates beyond the canonical golden-section deviation. We list four single-experiment falsifiers, including a precision measurement of the ³He(α,γ)⁷Be astrophysical S-factor at BBN energies.