I don't really understand the point of differentiating the wins and losses with a formula-based metric after you've already broken it down to individual seasons. Feels like he's trying to reduce sample sizes for the sake of spurring discussion.
One thing that I found interesting was looking at losses with a 36+ Game Score and subsequently winning % in those games:
It probably doesn't mean anything but it's interesting that Magic's team never lost a regular season game when his Game Score was 36+ (at least since the 83-84 season) and Westbrook's team is somehow close to 50%.
Also, in the interest of increasing sample size here are the team records of players who have had at least 4+ playoff games with a Game Score of 36+: