Agreed. I think this kind of research can give ideas on what kind of experiments to run on pricing pages, but there are so many factors affecting real life plan choice that everyone should have their own A/B testing setup to see what works for them specifically.