We can agree to disagree. My claim is actually quite the reverse. For A/B testing specifically, Bayes is much better suited to address the practical questions you would usually have when running A/B experiments. See my response to AlexeyMK below.