We benchmark five microscopic cuprate superconductivity theories against ten experimental constraints using continuous figures of merit $f_i\in[0,1]$ with explicit acceptance windows.  At mean-field level, Model~B (SVILC) scores highest ($F=0.96\pm0.02$), followed by Model~E (spin fluctuation, $0.91\pm0.03$).  Failure analysis identifies four ingredients that appear necessary at mean-field level: Mott physics, spin-fluctuation pairing, loop-current order, and three-band CuO$_2$ structure.  A unified Hamiltonian incorporating all four yields $\Delta_0=28$~meV and $\mu_{\mathrm{orb}}=0.06\;\mu_{ B }$/Cu on a $24\times24$ cluster.  The multi-observable scoring framework provides concrete targets for quantum simulation of cuprate Hamiltonians and enables systematic validation of quantum-computed phase diagrams. Four falsifiable predictions are derived.