After subtracting our CMB-frame motion, the Hubble flow still carries a dipole: a directional asymmetry in the apparent expansion rate persisting to large distances, imperfectly aligned with the CMB dipole direction (Secrest 2021; Singal 2022). FLRW expansion is direction-blind by construction, and a misaligned residual means two directional ingredients where the model permits one.
The model assumes one global expansion rate per epoch and exactly one directional contaminant, our peculiar motion, fully removable by one boost. Whatever survives that subtraction has no permitted source: no anisotropic expansion, no gigaparsec bulk flows, no second axis. The residual dipole is therefore homeless by construction.
SCT supplies exactly the second ingredient the data demand. The first axis is kinematic: our patch's residual frame velocity within the parent frame (P63), the collision-geometry memory the CMB dipole reads. The second axis is dynamical: the Λ_eff field is not isotropic around us, its gradient shaped by the KBC underdensity and the parent-frame mesh structure (P17, P19), so the apparent expansion rate genuinely differs with direction at the registered level, a dipole-like H₀ modulation of order 0.2 percent aligned with the local bulk-flow geometry. Two ingredients with different directions combine vectorially, which is precisely why the residual dipole sits near, but not on, the CMB dipole direction; the misalignment that embarrasses the one-boost model is the SCT signature.
The geometric bookkeeping is registered: the frame velocity runs along the impact parameter while the deposited J axis stands perpendicular (P64), tying this anomaly to the same two-axis anatomy carried by the count dipole (recid 77), the kSZ and ISW templates (recid 69), and the bulk-flow excess (recid 87). Sibling pockets at gigaparsec separations add the slow third term (P58, P59). The expansion-side machinery is laid out in Paper 7, From Chaos To Cosmic Expansion, with the collision kinematics in Paper 1, From Chaos to Convergent Foundations.
Keystone economy: P63 and P19 each carry one axis. The anomaly is not that a dipole survives subtraction; it is that the model only brought one direction to a two-direction problem.
The kill is registered against the all-sky supernova programs: LSST finding directional H₀ isotropic at 0.1 percent precision in every sky direction after dust and peculiar-velocity corrections rules out the dipolar Λ_eff variation at high significance and removes the M5 axis. The kinematic axis is separately exposed: a residual dipole that tracks the CMB dipole exactly, with no second component, would leave nothing for the Λ_eff gradient to explain and reduce SCT's two-ingredient anatomy to an unnecessary complication.