We present a hardened phenomenological audit of a primordial log-periodic feature program formulated within the Fractal Consistency Law and the Principle of Minimum Inconsistency (FCL-PMI). The central methodological question is whether the framework admits a rigid, low-complexity realization capable of producing a positive local fit improvement without relying on free feature-parameter tuning in the final run. The analysis formalizes a canonical family of primordial events, derives an observable anchor scale, compares three competing amplitude laws (B1, B2, and B3), and follows explicitly the reduction chain 2P to 1P to 0P. Within the audited family, the leading effective transfer law is the B3 rational-saturation prescription, which is the only tested law that simultaneously avoids over-suppression, remains bounded, recovers the weak-load regime smoothly, and preserves a genuinely zero-extra-parameter final realization. The corresponding primordial spectrum is implemented as a Gaussian-envelope log-periodic modulation of the scalar curvature power spectrum. The updated benchmark is materially stronger than the earlier local result. After local cosmological refitting, the internally favored Planck realization reaches a provisional improvement of approximately Delta chi-squared near 0.94 at the current working transfer width. Additional independent support is obtained from high-ell TT likelihoods in both plik and plik-lite. Proxy external portability checks of the cosmological displacement induced by the feature-favored solution are positive in ACT DR6 and in ACT combined with low-ell EEsroll2, while remaining mildly unfavorable in SPT TEEE. The resulting evidential profile is therefore favorable but not uniform: strongest in temperature-dominated sectors and mixed in polarization-dominated checks. The signal remains weak and does not constitute robust evidence for new primordial physics. Its relevance is methodological. It shows that a theoretically fixed, low-complexity structure can remain phenomenologically competitive after refitting, can survive internal temperature-sector decomposition, and can exhibit partial external support without recourse to final-run feature tuning. The paper therefore upgrades the status of the primordial-feature block of the FCL-PMI program from a merely suggestive architecture to a disciplined, externally probed viability benchmark.
César Daniel Reyna Ugarriza (Thu,) studied this question.