I hesitate to get too deeply into "what is evidence" with you, but suffice it to say that evidence for a proposition is that which renders the proposition more likely.
Comparing Bolt's current 100m season (9.88 +1.0,[9.93] 9.91 +0.6 [9.94], and 9.91 -0.2 [9.89]) to either his 2008 or 2009 100m seasons, and drawing conclusions from that comparison, is entirely legitimate.
The conclusion you have drawn, based on that comparison likely combined with other evidence, is that "he was doping in the past". That is precisely the conclusion I reached, using evidence that I used, in conjunction with other evidence.
Some of the other evidence I used is, for example, that a 9.58 IS too fast. While I build a case to show the strength of that evidence, based on such things as historical performance, I add to the strength of that evidence.
Your suggestion that Bolt may be a 1 in 100 years freak
is not evidence of the contrary proposition--that he was clean--but instead would tend to reduce the WEIGHT of my evidence. The degree to which it reduces it is related to the quality of your evidence, for which there is none--it is a mere suggestion, not a fact. The only actual support for your suggestion would be the existence of the 9.58, the very performance in question. For your suggestion to have any merit, you must assume what you set out to prove--that 9.58 was a clean time, and that Bolt is therefore a 1 in 100 years freak.
Your suggestion, having to assume what it sets out to prove, is therefore worthless, and does not diminish the weight of my evidence at all.
You might not understand this precisely, but intuitively you know it to be true--that is why, even in view of your suggestion that he is a 1 in 100 years freak, you nevertheless reach the conclusion that he was doping in the past.
His corrected average 100m this year, when potentially clean, is around 9.91, which is certainly excellent from an historical 100m season perspective, but it doesn't seem like the work of a 1 in 100 years freak--I don't even think it's any better than, say, Powell's 2007 or 2006, or maybe even 2005 seasons. Two 1-in-100 years freaks at the same time, from the same country? Even less likely than your suggestion of one 1-in-10 years freak, which again has no support whatsoever.
Again, his 2008 or 2009 100m seasons are not evidence of his being a 1-in-100 years freak, because again that would assume what is to be proven.