Strong-lensing clusters spent a decade measuring 50 to 100 percent overconcentrated against ΛCDM's relation, suggesting cores formed earlier and denser than hierarchical growth permits, before the audit deflated the headline: mass-concentration anti-correlation biases fits steep, lens selection prefers elongated systems seen long-axis-on, and CLASH with proper selection modeling agreed with simulations at 90 percent confidence. Extreme individual lenses still sit high, and the relation's evolution remains untested.
Concentration is the model's formation-clock fossil: the relation and its scatter are fixed by assembly history, so any persistent overconcentration indicts the clock. The audit saved the relation by charging the excess to selection, leaving the model owing only the outliers and a prediction for the deeper census.
SCT reframes what the fitted concentration measures: lensing profiles are baryons times A(r), and forcing that shape into an NFW template returns an effective concentration set by the coherence profile, the registered concentration-averaged structure (the A_NFW bookkeeping at 5.83 against the fixed point 5.970), with the apparent concentration tracking coherence state rather than formation epoch alone. The audit's two findings then read naturally: agreement of the bias-corrected bulk with simulations is expected, since coherence-amplified profiles were calibrated to reproduce observed cluster masses, while the persistent extreme outliers are the high-coherence tail, systems whose dynamically cold, aligned ensembles (the elongated, long-axis lenses selection prefers) carry A above the average and lens accordingly. Strong-lens selection preferring overconcentrated systems is, in coherence terms, selection preferring coherence, the same bias the GGSL excess quantifies inside the arcs.
The framework's stake in the Euclid retest is directional: at fixed mass and selection, fitted concentrations should correlate with coherence diagnostics, alignment, ellipticity, dynamical coldness of the member ensemble, residual structure the formation-clock reading does not predict once selection is fixed in advance.
This is the same NFW-template bookkeeping behind the lensing-is-low and substructure entries, read in the concentration parameter. There is no need to choose between a broken clock and a closed case; the parameter was measuring something else.
The Euclid census carries the test: with pre-registered selection functions, fitted concentrations statistically independent of coherence diagnostics at fixed mass, the outlier tail dissolving into pure projection statistics, would remove the coherence reading and leave concentration a clean formation fossil. The registered cluster identities bound the other side: concentration-averaged amplification drifting from the A* bookkeeping across the census would break the underlying profile family.