Fascinating stats and thank you for it!
However, just to point out that it's not really accurate to line up a player's plus/minus against the team's overall goal differential because, of course, the player's plus/minus doesn't account for power-play goals (but the team's overall goal-differential does). Thus, there's a mis-match in the analysis, particularly in the case of high-scoring forwards and defence that would regularly be on the power-play.
For example, to take Gretzky's best on-ice goal-differential season (1984-85), he was +98. In that season, the Oilers scored 74 power play goals and had 76 scored against them. I would guess that Gretzky was on the ice for maybe 75% of their power play goals scored -- he had 43 power-play points (another +56) -- and was killing penalties for maybe 40% of the power-play goals against (another -30). Combined, then, his "actual" on-ice personal goal-differential is +98 plus maybe +26 = +124. That's the kind of number that should be aligned to the team's overall goal differential.
I suppose we can infer that high-scoring forwards on bad teams who don't kill penalties would have a significantly improved plus/minus if we could see their "actual" plus/minus, regardless of special teams. (In Gretzky's case, that would probably improve his early-seasons' plus/minuses a lot, but after about 1984 or so he starts killing penalties more, and thereafter it would improve his plus/minus less, but still somewhat.)