Peak elo is determined by your score against a pool of players. Average elo is a better indicator of strength.
Players such as Robskate (especially), Croboys, Lordy and Yuzgen achieved their peak elos in 2017-2018 when the ladder was relatively new and the big guns still hadn't started playing yet. This is fact. If you take their peak elo from 2020 (pandemic era) onwards, I imagine you would find they would rank outside the top 20.
The data is also false. Yuzgen has used 2 accounts on different computers (Ram-yuzgen) and (Ram-yuzgen_). If you combine both accounts I am certain he has played more than 15,000 games. He has played by far the most amount of games of anyone and this can be verified by fact by going through each month and adding up his total amount of games played. For example, I didn't start playing ladder until April 2020 during the Pandemic, Yuzgen had already been playing for 3 years before that.
There's also some obvious data input flaws as well. For example it says Maddux only played around 400 games in Diaz's list. If you look at April 2020 month, he played 500+ games alone (username: Rocket). Therefore the data is clearly not accurate.
Numerous fatal flaws with the data presented and I don't give this ranking any validity. Stand out example: Faust is far too low. Players like myself and Faust have always played in months where heavyweight ranked players have played (Ford, Dynamic, Beast for example), therefore there is probably a ceiling to the elo we could achieve without improving our play. Faust would obviously rank inside the top 7 at minimum, and probably top 5 overall. Robskate achieved a peak elo playing against noobs on standard maps back in 2017 in contrast. Not a single player on CnCNet would argue robskate is a better player than Faust, so this disproves the whole list right away.
On the whole, the data is meaningless. Go back to the drawing board Diaz and number crunch more accurately if it interests you. Also average elo would be a better indicator of strength if you're able to provide that (but only if you make sure all games are accounted for (eg Yuz's missing 5000+ games).
Players like Leigh with also such a small sample size of games shouldn't really count. Elo doesn't really mean anything until there's enough data size.
Also, where is Mclovin on the list ???? Clearly a data input flaw and August 2020 is missing (the month Mclovin won), and probably the month I myself achieved my peak elo in 2nd place and destroyed the rest of the competition.
Regards