A good question. I think that the evolution of World Records has more to do with the evolution of racing than with any new training.
Please bear in mind that these training methods are relativley new, i.e. since the early seventies, which is the time I believe when it was realized that the pioneering days were over, and 'modern training' was established. Since then all the best runners do variations of these types of training.
So that's less than 40 years with a steady progress of race times coming down, due to more money, better fields, more fast races and many more full time runners. Add to this, the fact that EPO was certainly being used by some runners, although not necessarily the fastest ones.
If you look back to the seventies, runners like Bill Rodgers did a huge number of races. I think in one year he did seven Marathons, which begs the question; how fast could he have run if he had today's advantages? Although he would probably answer that he loved his Marathons, and got a huge amount of public support.
So isn't it true to say that in a fast race, we should expect the fastest guys to be running 2.04-2.05 in the fastest races, and that when conditions are average, 2.07-2.08 is also to be expected?