Groups and clusters break self-similarity, L_X scaling as T to the 2.6-to-3 against the predicted T squared, because their cores carry an entropy floor of 100 to 300 keV cm² requiring 1 to 3 keV per particle of non-gravitational heat (Ponman 1999). The floor is universal across a factor of thirty in temperature, in active and silent systems alike, yet the model must build it stochastically, from AGN histories tuned to converge on one pedestal everywhere.
The model's gas starts cold and unprocessed: every joule above gravity must be injected by feedback inside each halo's own history. A universal pedestal then demands universal histories, which stochastic black-hole astrophysics does not supply, so the simulations purchase convergence with tuned duty cycles in every box.
SCT deposits the floor once, for everything: the intracluster and intragroup medium is collision-thermalized material whose entropy was set at the cascade and conserved adiabatically ever after, the relic K₀ with the registered quantitative chain tying the 100-to-300 keV cm² range to the collision energy parameter constrained to its narrow physical window (P22, P25, P29, P30). Universality stops being a burden and becomes the signature: one thermalization event stamps every parcel of proto-ICM with the same pedestal, independent of subsequent black-hole biography, which is why silent-nucleus groups carry the same floor as active ones and why the L_X-T slope steepens identically across the ladder, the registered effective-slope structure landing at the observed 2.6 to 3.
The registered prediction set makes the costume removable: the floor must be (a) universal across cluster masses, (b) correlated with large-scale structure density, (c) unchanged from z of 2 to now, and (d) present equally in AGN-quiet and AGN-active systems, four discriminants against any preheating account, whose floor assembles over time and tracks feedback history. The quiescence constraint comes free: a deposited floor heats nothing and stirs nothing, coexisting with the calm cores that embarrass distributed-feedback heating.
This is the keystone thermodynamic relic behind the cooling-flow stability, the quiet cores, and the born-hot protoclusters: one deposition, four X-ray observables. There is no need to manufacture initial conditions dynamically in every halo's past.
The registered kills are explicit: Athena-class spectroscopy finding K₀(z = 2) far below K₀(z = 0), the floor assembled late, falsifies the relic directly; AGN-quiet groups showing a shallower L_X-T slope than active ones certifies preheating and removes the deposit. The evolution measurement is the decisive one, and the relic reading requires the high-z floor to be already in place.