The scary part is that far too many people's attitudes seem to be "well, so what if he cheated in online games, everybody does that, it's meaningless...we're talking about OTB...".
Well that's certainly not my opinion. The only way that I view the earlier stuff as somewhat forgivable is that it was done when he was a very young kid. Not that he's not still very young but you know what I mean. I don't consider a 12 or 16 year old immune to the temptations of cheating. And I'm pretty sure that people can change quite a bit from their early years. Not to say they all do but they can. I'm just willing to give the benefit of the doubt until someone proves to me otherwise.
Let me point out the fundamental point that it is mistakes that change the true evaluation. Good moves don't. This parallels what happens with a powerful engine that anticipates the possibility of a "great move" (or just a good move) being played and its evaluation of that move merely becomes a little deeper after it has been played, so usually stays comparable in character.
Note also that the way a players' moves relate to computer moves when their evaluations are not so different provides more subtle evidence of cheating. Merely playing the moves that are clearly best and avoiding very bad moves is much weaker evidence in general.
It's probably also worth pointing out that while the analysis tool provides information that is interesting, it is a million miles from the sort of analysis that is used to detect cheats that are not really blatant.
Fair enough, but it should also be noted that these blue great moves did not follow any significant mistakes. The graph remained basically a flat line. You can look at the evaluation graph from the game review link I posted above. But yes Point well taken, good moves don't increase the eval very much.