The baryon-to-photon ratio, eta = 6.1 × 10⁻¹⁰ (baryon loading R_b near 0.26), anchors the abundances, the acoustic peaks, and the matter budget, measured to a third of a percent, and ΛCDM cannot compute it: baryogenesis is delegated to beyond-Standard-Model mechanisms with no laboratory support and enough freedom to fit any value. Cosmology's most precisely measured number has no predicted value in the model that depends on it.
The model takes eta as an input by construction: a singular origin offers no structure from which a pure number could follow, so the ratio is whatever unknown asymmetry physics happened to produce. The more precisely it is measured, the more conspicuous the absence of any derivation becomes.
SCT computes the number. The collision cascade partitions energy through SO(3) geometric channels (three cascade modes), loses a derived 13.6 percent fraction at the QCD boundary crossing, and heats the photon bath by a derived 2.3 percent: the chain yields R_b = 0.2545 +/- 0.032 with no BBN or CMB input anywhere in the derivation (Paper 17's cascade-geometry result, built on P36, P37, P38, P39). The observed value 0.260 +/- 0.002 sits 0.17σ from the derivation: agreement at the level the standard model reserves for fitted parameters, achieved by counting geometric channels. The most precisely measured free number in cosmology stops being free.
The derivation's honesty matters: it closes the circularity objection to earlier SCT work that had matched R_b from observation, because the geometric chain runs forward from collision structure alone, making the 0.17σ agreement a test passed rather than a calibration absorbed. Baryogenesis proper, the matter-antimatter asymmetry, runs through the companion geometric channel (P41), with the collision J vector supplying the CP-breaking axis; the two derivations share one cascade.
This is the same cascade geometry that yields N_eff = 2.514 and the CAR sound speed: three numbers from one structure, none adjustable. There is no need to invoke unknown asymmetry mechanisms optimized for viability.
The registered kill: the observed R_b falling outside the derivation's 3σ window of 0.158 to 0.350 falsifies the cascade-geometry chain; within the present precision the test is passed at 0.17σ, and the bound tightens as the derivation's uncertainty (dominated by the QCD-boundary loss term) is refined. The companion predictions share the jeopardy: N_eff measured at the Standard Model value of 3.046 by CMB-S4 at 3σ against 2.514 would break the same cascade structure that delivers R_b.