This Looks Better than That: Better Interpretable Models with ProtoPNeXt