A 4 hour stage event is well below your VO2max (an 11 minute all out effort) -- maybe you can explain how oxygen delivery is always a weakness? In any case, I used the term first, so if you used it differently, ...
Basically, a drug only works, if it's addressing a weakness. The benefit is relative to the amount that weakness can be strengthened. This is why it might help a 15:00 5K runner more than a 13:00 5K runner -- because of the difference of their state of training. This is what I meant by weakness.
I don't say that EPO doesn't work at all. Clearly studies show that it works for running and cycling. (But I never heard of a study on elite athletes.) I say that the benefit depends on the athlete, and his current state of fitness, not to mention detailed factors like dosage and timing. It would be a mistake to cling to a percentage improvement, or an absolute time advantage. EPO might give one person 20 second advantage, another 5 seconds, another 1 second, and another becomes sick. In different circumstances (changing dosage, frequency, training, ...), the same people might get bigger or smaller advantages.
From a scientific point of view, it is risky to draw conclusions from one sample, and apply them to a different sample. 18:00 runners are not the same as 13:00 runners. Cycling 2000 miles in 3 weeks is not the same as running 26 miles in 2 hours. Extrapolation and projection are useful scientific tools, but conclusions can only be drawn understanding that risk.
How do I think people misunderstand EPO? People act as if it's a light switch, that provides some predictable percentage, or absolute time, of improvement. "Everybody doped, therefore the playing field was level". It's also more than about improving "oxygen delivery" during a race. It's also effective during training, and after a race, helping you to recover faster. While EPO is helping peak power on climbs, it is also helping cyclists push harder in stages, and not bonk the day after. This is an example of a benefit for cycling grand tours, that doesn't translate at all to running a 10K or a marathon -- there is no race the next day. But then again, steroids, HGH, and testosterone also provide both strength and recovery. The contribution of EPO is only one part of this multi-cocktail doping regime.
I'm not sure what you think my opinion is. My opinion is to be reasonable. My opinion is that cycling 3 week grand tours, is not the same as running a 10K or a marathon. My opinion is that a drug that provided "global" benefits to cycling cannot provide "regional" or "local" benefits to a few tribes in poor nations. My opinion is don't jump to conclusions, but use supporting facts to build bridges, or stepping stones, and take baby steps. I think it's naive to leap to conclusions, yet I'm often called naive for not making that leap entirely on faith.
The very premise of this whole thread -- that Geb and Bekele are equally guilty as Lance -- is a clear example. It accuses Lance haters of applying a double standard. The same rigid standard allows me discredit Lance, while defending Geb, Bekele, Komen, Lagat, and even El Guerrouj. I need more supporting facts than coincidence in time, and incredible times. There are a dozens, if not hundreds, of known facts against Lance that don't exist for Geb and Bekele. They are not equivalent.
Regarding EPO effectiveness, what kind of proof can I, or anyone else, provide? I know there are EPO studies on good, amateur cyclists. I know there are EPO studies showing 40 second improvements over 10K (for the subjects in the study). I also know that "hi-lo" altitude training also gave a 20 second improvement for subjects over a 5K. There's some proof that I already offered that makes me ask, if sea-level subjects can adapt to altitude in 4 weeks, and maintain the effect for 4 more weeks, why wouldn't East Africans already have some of this oxygen delivery benefit? And my unsupported, but I think logical, question is, wouldn't that make EPO less effective for them (since they have already strengthened that weakness)?