Bridgeburner: <Dr. Funkenstein>
<There is also no attempt to analyze how often a 2700 player (basically the level that Ivanov played at in this tournament) follows computer lines in a game exactly and comparing Ivanov's % to it.>
A did so a couple of years ago when engine mapping and analysing the games of Lasker and Schlechter's 1910 World Championship match and the 2008 World Championship match between Kramnik and Anand.
I started by summarising the match between player moves and engine preferences, but stopped doing it when I concluded that was an unnecessary detail. However, every move of every game in these two matches was subject to exhaustive engine analysis, and mentioned wherever an engine had a preferred move. These can simply be counted up for each game.
It's all accessible from here: User: game mapping project
For my own part, I have no hesitation in accepting Lilov's conclusions.