We operationalize residual structure in the mod-6±1 support space U6 through bad sets, shift indices, and masked carrier/filament ablations. The contribution of this paper is deliberately limited. GL1 III is neither a formal GL (1) bridge nor an RH claim. It is the empirical and protocol-facing intermediate layer between GL1 II (the U6 geometry and composite explicit formula) and GL1 IV (the T5-replacement / lift program). Boundary- Ladder measurements provide evidence-bearing indications of (i) finite carrier-capture depths at selected hot boundaries, (ii) reproducible shift regimes under seed, resolution, and halfwidth variation, (iii) non-degenerate bad masses and guard stability, and (iv) a baseline null separation against permutebins. This paper does not claim the stronger pushforward covariance of carrier profiles or the lift into the operator-side diagonal metric; both knots remain the subject of GL1 IV. In this form GL1 III serves as a decision-capable evidence layer: it clarifies which arithmetic regularities would have to support a later transport lemma, without already promoting them to a formal lemma.
Stephen Steiner (2026) studied this question.