Cluster member galaxies produce galaxy-galaxy strong lensing at rates exceeding simulations by over an order of magnitude (Meneghetti 2020, Science), narrowed to a stubborn factor 2 to 4 with full baryonic physics (Ragagnin 2022), with observed Einstein radii of 2 to 5 arcseconds against simulated 0.3 to 1. No CDM mechanism supplies the compactness, SIDM moves the wrong way, and the model lands in an inverted crisis: too much substructure by count, too little by lensing punch.
Member-galaxy lensing power is fixed by subhalo density profiles that collisionless collapse plus tidal processing determine: the simulations are the prediction. An order-of-magnitude rate excess means the real substructures are far more gravitationally concentrated than CDM can make them, and the model has no compactness dial left to turn.
SCT supplies the missing compactness without rebuilding any subhalo: cluster member groups are coherent comoving ensembles whose effective lensing mass carries the substructure-scale amplification, A_sub of 3.2 to 4.0 from the registered coherence chain, and lensing cross-sections scale as its square, the factor 10 to 16 that converts simulated GGSL rates into observed ones (P50, P52, P53). The registered amplification-ratio structure ties the substructure boost to the cluster-scale value through the measured velocity dispersions and sizes of member groups, so the same bookkeeping that gives clusters A* at the virial radius gives their compact cold subgroups the stronger boost the arcs record. Einstein radii inflate accordingly: theta_E scales with effective mass, and amplified members lens like the compact objects the simulations cannot manufacture.
The inverted crisis becomes diagnostic: counts follow the visible debris census (modest, as the satellite entries document) while lensing punch follows coherence, so SCT predicts exactly the observed signature, fewer-but-punchier, where CDM predicts many-but-soft. The registered discriminant survives: the excess should scale with member-group coldness and compactness, strongest for dynamically cold, tight subgroups, measurable across the growing JWST census.
This is the same A_sub structure behind the flux-ratio anomalies, audited at its cleanest scale. There is no need to demand compactness from collisionless collapse that tidal physics forbids.
The registered kill: if uniformly increasing CDM subhalo concentrations within simulations fully resolves the GGSL excess at all radii and masses without violating field constraints, the excess is a concentration artifact and the amplification is unnecessary. The scaling cuts independently: GGSL rates statistically independent of member-group velocity dispersion and compactness, the coherence diagnostics, would break the A_sub account while leaving the raw excess for someone else to explain.