This is solid, but the experimentation bit at the end is missing one crucial point — you need volume to experiment. Any pricing study, whether of traditional elasticity or segment behavior, needs a fair amount of volume to be accurate. If you have a high priced, lower volume thing, say 5 figure enterprise software, be very judicious about experimenting. The results are less likely to be statistically significant and complex multi-stakeholder buying processes (say through purchasing) can’t be tested in an email.