The radio sky is crossed by enormous coherent arcs, and the most ambitious measurement in cosmology must see through them. Loop I, the brightest, spans over 100 degrees from the Galactic plane toward the north polar cap; a dozen catalogued loops and spurs cover much of the high-latitude sky in polarized synchrotron emission. Their standard biography makes them old, nearby supernova remnants or wind bubbles, but the accounting has never closed: distances are contested between tens and hundreds of parsecs, some loops trace coherent magnetic structures that ordinary remnant evolution does not produce at the claimed sizes, and polarization analyses show loop-correlated features extending to high latitudes exactly where CMB polarization treats the sky as clean. Planck-era studies demonstrated that Loop I's polarized emission is detectable in the cosmological channels, and dedicated analyses have warned that loop residuals could mimic or contaminate B-mode signatures, the same class of foreground misjudgment that converted BICEP2's primordial claim into a dust measurement.
For ΛCDM the loops are local nuisance, but the nuisance is structural: foreground cleaning relies on spectral templates whose validity assumes the synchrotron sky is statistically well-behaved, while the loops are singular coherent objects with their own spectral indices, polarization angles, and possible high-latitude extensions that template methods are not built for. An unmodeled coherent foreground arc is precisely the failure mode that produces false primordial detections at the sensitivity of LiteBIRD and CMB-S4.
The standing is a known unknown of the B-mode era: the loops are real, bright, polarized, badly understood, and draped across the cleanest patches of sky, and the primordial gravitational-wave program inherits whatever errors their modeling carries.