At the largest angular scales the cosmic microwave background carries less power than ΛCDM predicts. The deficit spans the multipoles below about 30, where observed power runs roughly 60 to 90 percent of the best-fit expectation depending on the bin, and it is sharpest at the very largest scales: the quadrupole comes in at roughly a quarter of its predicted value. The feature has persisted across three independent experiments, COBE, WMAP, and Planck, surviving every change of instrument, scan strategy, foreground treatment, and likelihood, which makes an instrumental origin hard to sustain. It travels with a family of companions at the same scales: the quadrupole-octupole alignment, the hemispherical power asymmetry, and the parity asymmetry of even versus odd multipoles.
The structural problem is that ΛCDM with inflation predicts the opposite of a cutoff. Inflation stretches fluctuations to arbitrarily large scales, populating modes far beyond the horizon, so the primordial spectrum should continue smoothly through the lowest multipoles with no preferred scale. Within the model the only defenses are cosmic variance, since few independent modes exist at low ell, the individual deficit being a roughly 2 sigma accident, or bespoke modifications: a feature in the inflaton potential, a pre-inflationary fast-roll phase, or late-time ISW compensation, each engineered for the symptom and none independently motivated. Cosmic variance can excuse the deficit alone, but the simultaneous alignment, asymmetry, and parity companions at exactly the same scales strain the accident reading.
The standing is steady: the deficit itself is modest in significance and will never improve much, because the largest scales have been measured to cosmic-variance limits in temperature. What can move is the joint statistics: polarization at low ell (LiteBIRD, CMB-S4) provides nearly independent modes over the same scales, capable of confirming whether the large-angle sky is genuinely anomalous or merely unlucky.