You mentioned OPAmp, not an OTA before an inverter or comparator. If you say OPAmp I think most of us will assume negative feedback, which isn't possible with a comparator.
Comparator delay simulation/measurement is not equal with slew-rate. If you know that only large differential signals will appear you can suppose the slew-rate determines the worst delay. Otherwise you have to know the minimum input overdrive voltage and the delay will be much larger with that.