PUH Theorem 354 (The Constraint Does Not Reach the Prediction) — ACT DR6 Reconstructs Anisotropic Birefringence from L = 15 Upward Against a Scale-Invariant Axion Template, While a Rebound Axis Is a Dipole at L = 1 and a Global Handedness Is the Monopole the Analysis Subtracts as a Systematic

Abstract

Photonic Universe Hypothesis (PUH) — Positioning Note. THE MEASUREMENT. Kramer et al. (arXiv:2608.27458) constrain two secondary CMB anisotropies with ACT DR6: anisotropic screening from Thomson scattering off free electrons, and cosmic birefringence, the rotation of CMB polarisation by parity-violating physics beyond the standard model. BOTH ARE NULL. Anisotropic birefringence: A_CB = 0.05 ± 0.06, upper limit 0.15 at 95% (conservative 0.52), corresponding to a Chern-Simons bound g_aγ < 4.9×10⁻²/H_I. Anisotropic screening: A_τ = 180 ± 157, upper limit 450 (conservative 370). TWO CONTEXTUAL POINTS THE ABSTRACT DOES NOT EMPHASISE: the birefringence limit is WEAKER than earlier work (0.021 from Planck, 0.044 from BICEP/Keck) because of polarisation noise and the ℓmin = 600 restriction; and the screening analysis finds foregrounds severe enough that the CIB-deprojected consistency test FAILS OUTRIGHT at PTE < 10⁻⁵. RESULT 354.1 (the prediction lies below the measurement). T216 predicts that particle-scale chirality and cosmic-scale rotation share one axis — the rebound polar axis. AN AXIS IS A DIPOLE, AND A DIPOLE IS L = 1. The reconstruction runs from L = 15 (baseline) or L = 40 (conservative) to L = 3000. SO THE AXIS PREDICTION LIES ENTIRELY BELOW THE LOWEST SCALE MEASURED, by more than an order of magnitude in multipole. AND THE MONOPOLE IS WORSE THAN UNMEASURED: the analysis states that a uniform rotation angle contributes to the monopole of the birefringence spectrum and SUBTRACTS IT AS A MEAN-FIELD BIAS, using the instrument's global polarisation angle of 0.2 ± 0.1 degrees. A GLOBAL PARITY VIOLATION OF EXACTLY THE KIND A REBOUND HANDEDNESS PRODUCES IS ELIMINATED BY CONSTRUCTION BEFORE THE MEASUREMENT BEGINS. RESULT 354.2 (the template is axion-shaped). The constrained amplitude is defined against C_L^αα = A_CB · 2π/[L(L+1)] · 10⁻⁴ rad² — a SCALE-INVARIANT spectrum motivated by a nearly-massless axion during inflation, with the amplitude tied to a Chern-Simons coupling between the axion and the electromagnetic field strength. THIS FRAMEWORK CONTAINS NO AXION AND NO SUCH COUPLING. Its source is the rebound's handedness; its spectral shape is a dipole. FITTING IT TO THAT TEMPLATE WOULD BE A CATEGORY ERROR, AND THE g_aγ BOUND DOES NOT CONSTRAIN IT. This should be stated before someone else states the reverse. RESULT 354.3 (a discriminator the paper supplies unintentionally). The analysis distinguishes two possibilities: a pseudo-scalar distributed uniformly like dark energy gives an ISOTROPIC effect; one interacting through gravity is ANISOTROPIC and TRACES THE MATTER OVERDENSITIES. THIS FRAMEWORK FALLS ON NEITHER TINE — its parity violation is aligned with a single axis fixed at the origin event, tracing neither the matter distribution nor nothing at all. That is a THIRD signature and it is distinguishable from both: a matter-tracing signal correlates with galaxy density, an axis-aligned signal does not, and astrophysical backgrounds track structure — so A DIPOLE MISALIGNED WITH THE MATTER DISTRIBUTION IS HARDER TO FAKE THAN AN EXCESS. RESULT 354.4 (the instrument exists even though the test has not been run). The analysis supplies a working quadratic estimator for the birefringence field, built on an established lensing pipeline with the disconnected bias, secondary lensing bias, mask transfer function and Monte Carlo residual all characterised and removed — AND A PUBLISHED RECONSTRUCTED MAP over roughly 9,400 square degrees. SO THE TEST THIS FRAMEWORK NEEDS IS TO TAKE THAT MAP AND ASK WHETHER ITS DIPOLE ALIGNS WITH THE COSMIC FILAMENT SPIN AXIS. That is a different analysis on the same data and IT HAS NOT BEEN RUN BY ANYONE. And it is a CORRELATION rather than a detection, which matters because T216 predicts a direction and supplies no amplitude — a shared axis can be tested without knowing how large the signal is. TWO NOTES. ON THE PERSISTENT ISOTROPIC HINT: a nonzero uniform rotation angle of 0.35 ± 0.14 degrees (Minami & Komatsu 2020) has survived several reanalyses. IT IS NOT EVIDENCE FOR THIS FRAMEWORK and is equally consistent with other explanations. But it is the one existing signal whose SHAPE matches a global handedness fixed at the rebound, and the analysis considered here treats it as an instrumental systematic to be removed. ON A DISTINCTION THAT MUST NOT BE BLURRED: the vacuum birefringence treated in T321 is a QED effect in a magnetar's magnetic field; cosmic birefringence is polarisation rotation over cosmological distances from a parity-violating coupling. THE TWO SHARE A WORD AND NOTHING ELSE, and conflating them would misattribute a constraint from one to the other. KILL-CONDITIONS: (i) IF THE REBOUND PRODUCES A SCALE-INVARIANT RATHER THAN DIPOLAR PATTERN, the axion template applies after all and the published limits do constrain this framework — T216 predicts an axis but DOES NOT DERIVE THE MULTIPOLE STRUCTURE of the resulting rotation field, and that derivation has not been done; (ii) if the dipole of the published map is inconsistent with the filament-spin axis, Result 354.4's test fails and T216 with it; (iii) if the isotropic hint is confirmed instrumental, the one shape-matching signal disappears; (iv) if a future analysis reconstructs L = 1 and finds no dipole at any amplitude, the prediction is excluded. NOT CLAIMED: that the framework predicts cosmic birefringence at any particular amplitude, since T216 supplies a direction and no magnitude; that the isotropic hint supports the framework; that the published analysis is deficient — IT IS A CAREFUL PIECE OF WORK whose scope simply does not include L = 1; that the dipole test has been performed; or that this note constitutes evidence of any kind. IT ESTABLISHES ONLY THAT A CONSTRAINT FREQUENTLY CITED AS EXCLUDING PARITY-VIOLATING MODELS DOES NOT REACH THIS ONE.