https://pinkerton.com/our-insights/blog/age-crime-curve
(Yes yes I know, Pinkerton are evil. they have the best plot of this correlation I could find. The crime-age correlation is the strongest that exists in the entire field of criminology.)
If anything someone who is already known to cheat “just because” is even more likely to cheat when there is something to gain.
As for cheating and stakes I think it all depends. His claim is he cheated when he was 16 to boost his rating so he could player higher level opponents on stream and boost his career. If you accept that claim it would make sense that he rationalized it that he was just cheating to get to his "true" Elo and stopped cheating once he got there. Now Chess.Com seems to believe that he cheated beyond that but they haven't specified more at this point.
I see the pattern forming. He clearly has improved his play since but he could also have improved the cheating technique, as others pointed out, just needing a hint or two in the most decisive moments of the game. Has he not cheated against Magnus it's a pity that he got accused with no proofs.