Planck 2018 found the CMB acoustic peaks smoothed as if lensed 18 percent more strongly than the best-fit model predicts: A_lens = 1.180 +/- 0.065, more than 2.5σ above the value of exactly 1 the model requires. The same excess drives the closed-universe preference and splits parameter estimates between angular-scale ranges, while direct lensing reconstruction from the four-point function sees no excess at all.
ΛCDM assumes the only thing that smooths the acoustic peaks is gravitational lensing by the catalogued matter distribution. A_lens is not a physical parameter but a consistency check on that assumption, and the model offers no mechanism by which it can differ from 1. When the smoothing exceeds what the matter supplies, the model has nowhere to put the excess.
SCT replaces the hot-dense-center with a superluminal collision, and replaces the lone matter potential with Φ_eff(r) = Φ_local(r) + Φ_mesh(r) (P51): a coherent mesh contribution from the parent-frame hierarchy that adds real lensing convergence not attributable to catalogued matter (P50, P52, P53). Integrating the superposition boundary condition S(z₀) of about 4.4 over the CMB lensing kernel yields A_lens of about 1.18 (Paper 3, Section VII.3), and the independent coherence formalism of Paper 13 gives A_lens = 1.17 +/- 0.05. The PR3 anomaly was, on this reading, the mesh announcing itself.
The estimator split is informative rather than embarrassing here: peak smoothing integrates every gravitational deflection along the line of sight, while reconstruction templates are matched to the statistics of lensing by clustered matter, so a smooth coherent mesh component contributes more readily to the first than to the second. One secondary modulation layers on top: the dynamical Λ_eff (P17) shifts the background geometry slightly, which is why the same data pulled toward Ω_K < 0 when curvature was the only free knob.
This is the same mesh signature that appears as the closed-universe preference, the S₈ deficit, and the growth-index excess: one coherent contribution read through four parameterizations. There is no need to invoke positive curvature, modified gravity, or extra exotic clustering.
The anomaly has weakened in reanalysis: PR4/NPIPE gives A_lens between 1.04 and 1.10, ACT DR6 gives 1.007 +/- 0.057, and the PR4 reconstruction amplitude is 1.004 +/- 0.024. SCT's predicted 1.17 +/- 0.05 matched PR3 but now sits 1 to 2σ above the newer likelihoods; the mesh-lensing prediction is under live observational pressure and CMB-S4 will decide it.
The registered kill criterion: future CMB lensing measurements (CMB-S4, Simons Observatory) converging to A_lens = 1.000 +/- 0.005 would rule out any coherent mesh lensing contribution at the predicted level and falsify this mechanism outright. Conversely, a persistent percent-level excess in smoothing but not in reconstruction is the specific signature a smooth coherent mesh predicts and clustered matter cannot mimic.