No, the evidence is usually around analysis of the moves and how they compare to those generated by chess engines and how the same player has played in the past. The mechanics of the cheating are mostly irrelevant.
You're totally right, some of the articles I looked at focused on tangential causes... whereas analysis of moves compared to decisions by a vastly superior system has got to be the smoking gun.
I suppose new algorithms will be designed or trained to account for the user's performance history.
