Research route #48 · investment progress is separate from claim verification.
Route 48 pursue: Pascadi reading answered — the modulus is good, the length below c^{5/12} is the hard gap
Calibration: heuristic (a literature reading of Pascadi arXiv:2511.08445v2; the cited facts are from its abstract, §1.2 Theorems 1.1–1.2, and Example 1.3).
Result
I read Pascadi 2511.08445v2 to answer the four decisive questions the next experiment posed.
- Q1 (where the length enters). The method improves the trivial bound min(c, √(MNc)) only in the range M ≍ N ∈ [c^{5/12+ε}, c^{5/8−ε}] (Example 1.3, for c = p² or pq); Theorem 1.1 is stated for M,N ≪ c^{1/2+o(1)} with saving c^{−1/700}. The amplification is tuned to the M≈N≈√c balance point where the Weil and Fourier trivial bounds meet; below c^{5/12} the Weil bound √(MNc) is already the trivial bound and the method does not improve it. There is no sub-1/2 (or sub-5/12) extension in the paper.
- Q2 (near-prime exclusion). "Near-prime" is a modulus with a prime factor larger than c^{1−ε} (Theorem 1.2's condition — then no factorization c = dd′e with d ∈ (c^{1/3+ε}, c^{1−ε}) exists). The cell's modulus n = de is y-smooth, so every prime factor is ≤ y = c^{o(1)}: it is not near-prime. This is a positive sharpening — the modulus is a good (factorable) case for Pascadi, so the modulus is not the obstruction.
- Q3 (trilinear extension). The paper has no third-moment/trilinear extension; a trilinear second moment moving the exponent below 5/12 would be new work.
- Q4 (determinant-divisor coefficient). Pascadi's object is the fixed-modulus S(m,n;c) (S-family). The #935 determinant-divisor coefficient comes from the F-family phase e(−2ν inv(d)/e) with the denominator a summation variable; the non-abelian Fourier expansion does not touch that moving-modulus coefficient.
Conclusion
The reading is now precise: the blocking cell's failure is a pure length failure. The modulus n = de (y-smooth, factorable) is inside Pascadi's good set, but the cell's length n^{0.21} lies far below the improving-range floor c^{5/12} (≈ c^{0.417}), which is the lowest length any 2026 S-family method reaches. The experiment's success criterion (a source-matched mechanism at N < c^{1/2}) is not met; the failure criterion is only half met (the length barrier is real, but the near-prime exclusion does not cover the modulus). The remaining unexamined candidate is the iterated q-van der Corput method for smooth square-free moduli (Pascadi refs [55],[53]; origin [20], Blomer–Milićević [6]) — the one S-family line tailored to smooth moduli, whose improving range has not yet been checked against n^{0.21}.
Declared premises of this result: return #934 (accepted), return #935 (accepted).