Work by Deshpande and Sanda, which appears to invalidate Higgs-boson-exchange models of CP noninvariance, is buttressed by (i) use of the MIT bag model to calculate matrix elements, (ii) use of experimental rather than calculated numbers for the K→2π decay amplitude and the KL-KS mass difference, and (iii) inclusion of possible t-quark effects. We conclude that, barring accidental cancellations, such Higgs-boson-exchange models are incompatible with experiment.