Counting clusters should confirm the CMB, and instead it contradicts it. The abundance of massive galaxy clusters is exponentially sensitive to the amplitude of matter fluctuations, so cluster counts calibrated by X-ray and Sunyaev-Zeldovich mass estimates deliver an independent measurement of the cosmology Planck fixed from the primary CMB, and the counts come in low: fewer massive clusters than the Planck best-fit predicts, the cluster-count face of the S8 tension (Planck Collaboration 2015).
The designated suspect is hydrostatic mass bias: X-ray and SZ masses assume the cluster gas rests in hydrostatic equilibrium with the potential, and any non-thermal pressure support, turbulence, bulk motions, magnetic fields, makes the true mass larger than the hydrostatic estimate. The suspect, however, has an alibi problem: simulations and theory support a bias of 10 to 20 percent, weak-lensing calibrations of the same clusters constrain it below about 20 percent, yet reconciling the CMB cosmology with the observed counts requires a bias of 30 to 40 percent, well beyond what either the physics or the independent calibrations allow (von der Linden et al. 2014). The remaining options are new physics suppressing structure growth, or the admission that the masses, the counts, or the model are wrong somewhere unidentified.
The standing is the cluster-scale member of the growth-tension family, sharpening as eROSITA and SPT-3G counts grow and weak-lensing mass calibrations tighten the bias alibi further.