Agga, I had to look at this section because this paragraph looked weird to me. Here are three problems with it just on the surface:
If you go back and look at the ICC EHP thread, you'll see the run-time series statistical analysis of incoming hits, I think it was section 6. What matters in that analysis isn't the odds of taking a hit, but rather the odds of taking successive hits in a row at least one time druing a fight. At very low avoidance levels you're pretty much going to get hit a bunch of times im a row no matter what, as you increase avoidance levels they remain irrelevant until about 42% then got exponentially more important up until about 70-80% avoidance, then after that each increase becomes almost as irrelevant as it was until 42%.