It doesn't make much sense to translate the findings of this study on cyclists to the best times of elite runners.
Sand Dunes wrote:
That means on modern synthetic tracks and mircodoping. Peter Snell was capable of a 1:38 800m and 3:24 1500m.
It isn't scientifically defensible, no, but the back of the envelope calculations that Sand Dunes' is doing are the first things anyone should do when hearing about a small percentage performance increase and it's interesting that the results are fast but not otherworldly. Yes, 1:38 is fast for an 800m but in the right race with the perfect pacemaker, I think Rudisha could have approached that in his prime. Same thing if you convert Ron Clarke to 26:09. That's not completely outside what Bekele or Geb could have run in their primes.
But, but, rekrunner is an expert in exercise physiology and pharmacology, no-one knows better than him how these things add up.