MEON-R2 v1.21/v125 Consolidated Framework: Effective Reservoir Closure, Baryonic Torsion Activation, Conditional Occam Assessment, and Spin-Correlated Torsional Weak-Lensing Signatures

We present the MEON-R2 v1.21/v125 consolidated framework as an effective theoretical and numerical test construction combining R²/reservoir closure, an early-universe activation switch, spin/torsion anchoring, and the v125 baryonic torsion-activation scale. The v125 extension introduces the baryonic activation relation r_s(Mbar) = r_0(Mbar/M_0)^β with β ≈ 0.3, incorporated through the additional effective term Tbar/tors{μν}(r_s(M{ bar})) and source contribution Sbar(r,Mbar). In the v1.21 combination test, all internal numerical gates pass, including reservoir closure, balanced dark-sector bookkeeping, stable effective torsion evolution, externally anchored spin/torsion proxies, and CMB-compatible equality within the embedded comparison data. The preferred effective comparison yields Δχ^2=0 relative to the ΛCDM baseline in the included CMB test data. Under the conditional closure scenario in which the reservoir/baryonic structure replaces one effective free parameter, the framework shows an Occam advantage with Δ{ AIC}=-2.0 and Δ{ BIC}=-8.58; under a harsher count in which all R² hyperparameters are treated as free, the model is penalized. Additional supplementary toy tests examine the mechanical interpretation of the framework as a broad viscoelastic halo boundary layer, a stationary action-balance system with δ Seff≈0, a tensor-flow source-sector structure, a universal-circuit reservoir model, and a boundary-layer lensing analogue. The most distinctive proposed observational signature is not merely an excess halo amplitude, but a spin-correlated torsional weak-lensing residual: a very weak parity-odd or curl/twist-like component localized near r_s(Mbar), with a sign correlated with baryonic spin or jet handedness and suppressed under randomized-spin controls. This prediction is falsifiable and motivates future tests using spin-aligned weak-lensing stacks. These results should be interpreted conservatively. MEON-R2 v1.21/v125 demonstrates internal numerical consistency and a conditional Occam advantage in the tested effective CMB-compatible limit, and it motivates a distinctive torsional lensing search strategy. It does not constitute a final disproof of ΛCDM. Remaining requirements include a full covariant action-level derivation, official Planck high-/lensing likelihood evaluation, real SPARC-scale MCMC inference, and observational validation of the proposed spin-correlated weak-lensing signature.