Subtract our measured motion relative to the CMB rest frame, and the expansion should look identical in every direction; that is the cosmological principle expressed as a measurement. Yet analyses of recession velocities and distance indicators keep finding a residual dipole in the Hubble flow: a systematic directional asymmetry in the apparent expansion rate persisting to distances far beyond where local peculiar motions should matter (Secrest et al. 2021; Singal 2022).
The residual is doubly awkward. First, it should not exist at all: in an FLRW universe the expansion rate is one number per epoch, direction-blind by construction, so a persistent dipole requires either an unexplained bulk flow extending across the survey volume or a genuine anisotropy in the expansion itself, and the standard model can host neither. Second, the residual dipole's direction does not perfectly align with the CMB dipole, which undercuts the one escape route available, attributing everything to an imperfectly subtracted local motion. A misaligned residual means at least two directional ingredients are in play where the model permits exactly one.
The standing ties into the broader dipole controversy, alongside the source-count excess and the bulk-flow measurements, as part of a family of directional anomalies converging on the same uncomfortable question: whether the frame in which the universe expands isotropically actually exists. All-sky supernova samples from LSST and the maturing DESI peculiar-velocity program will measure directional H0 at the sub-percent level where the answer becomes unavoidable.