Distant TeV blazars appear harder than EBL absorption permits, an anomalous transparency claimed at up to 4σ (Horns and Meyer 2012) and disputed since, with axion-like particle conversion proposed to let photons tunnel past the background light. The accounting clashes internally: New Horizons suggests more diffuse optical background, deepening expected absorption, while the blazar spectra want less. The standard model's photon bookkeeping pulls against itself.
The model computes opacity from one homogeneous EBL: a single background density at each epoch, accumulated from its galaxy-formation history, traversed identically by every sightline. Any transparency anomaly must then be new particle physics, because the assumption leaves the background no room to vary between sightlines.
SCT breaks the homogeneous-EBL assumption the anomaly presses on: the background light is generated where structure is, and collision deposition concentrated structure into streams, walls, and the voids between them (P33, P34). EBL photon density is therefore inhomogeneous along sightlines, and TeV-selected blazars, preferentially detected along the cleanest paths, sample void-dominated corridors whose opacity sits below the homogeneous average: anomalous transparency without any new particle. The pull against the New Horizons excess resolves the same way, since the sky-averaged background (which the probe measures) and the path-selected opacity (which blazar spectra measure) are different moments of an inhomogeneous distribution and need not agree.
The framework's deep source ledger (P25, P46) adds the secondary layer: more faint emitters raise the mean background, consistent with the optical excess, while deposition geometry widens the variance, consistent with translucent corridors. The discriminating signature is then directional: transparency excesses should correlate with sightline void fraction, measurable by cross-referencing blazar spectra against cosmic-web reconstructions, where ALP conversion predicts magnetic-field-dependent but structure-independent transparency.
This is the same deposition geometry behind the void phenomenon and the homogeneity debate, applied to the photon background. There is no need to invoke axion-like particles to thin a background that was never uniform.
The directional test kills cleanly: CTA-era opacity measurements finding transparency anomalies uncorrelated with sightline void fraction, but correlated with intervening magnetic field strength in the ALP-predicted pattern, would falsify the inhomogeneous-EBL reading in favor of particle conversion. Equally, CTA finding no transparency anomaly at all retires the tension and the mechanism with it; SCT does not require the excess, only its structure if present.