Catwhoorg said:
I think 10W/Kg sustained for an hour would be clear proof of something being wrong, as would 9, 8 and 7. Much below that and the line gets fuzzy, and essentially becomes useless to consider.
OK, but "useless to consider" is another way of saying "I have no idea at all whether this rider is doping, therefore I will stay completely out of the discussion." As soon as one joins the discussion, and expresses an opinion about whether a rider is doping (
including commenting on other people's opinions on this matter), then one can no longer appeal to this agnostic view. It's like saying, I have no idea at all whether God exists, then criticizing people who argue that there is more evidence that God does not exist than evidence for existence. Unless your argument is that the evidence is exactly 50-50, such criticism amounts to an opinion, an interpretation of the fuzzy line.
Even fuzzy data can be interpreted in terms of probabilities, and they can be very useful even when unable to support certain conclusions regarding a specific rider. For example, if the probability of a certain power output being achieved clean is considered to be, say, 20%--based on the considerations that Ross discussed--then that certainly wouldn't justify a sanction. But beyond the fact that we could conclude the rider is far more likely doping than not, we could also conclude that if several riders all achieved this power level, the probability is beyond reasonable doubt that at least one of them is doping. This is basically the kind of logic that led people to look at the peloton a few years ago and conclude that there had to be doping.