Left: 3D radial distribution of the subhalos in our TNG-C cluster analogs (dashed, color-coded as indicated) compared to the force-disruption-free benchmark taken from P. F. Hopkins et al. (2023; black solid). We show the cumulative number of all subhalos with more than 20 DM particles, normalized to the total number within the respective 2R₂₀₀. To explicitly account for the artificial disruption of the inner subhalos, we artificially “inject” subhalos within 0.5R₂₀₀ by sampling weighted by the deficit within 0.5R₂₀₀ between the TNG-C and benchmark cumulative distributions (see the text for details). Three random examples of such modified subhalo radial distributions (dotted) are plotted that well match the benchmark distribution. Right: projected radial distribution of the subhalos hosting member galaxies in observed clusters (color-shaded, as in the top panel of Figure 2) and in simulated analogs with injected inner subhalos. The solid (dotted) lines show the mean (1σ confidence interval) inferred from 1 million bootstrapping iterations. Within the projected radius of ≲0.2R₂₀₀, individual radial bins remain 5σ–40σ discrepant with the observed abundance, even after explicitly accounting for numerical uncertainties due to artificial disruption.