This is really counter-intuitive. I would guess that the serve-receive asymmetry explains part of that (if players were allowed to serve 1 meter away from the net, all players would simply win 50% of the points, regardless of how good or bad they might be, apart from the cases where the player has a heart attack after the toss). Also, back to the historical greats, it would be interesting to see their "prime" stats, or at least the stats ignoring their early and final years. Even so I think it would be something like, say, 57% percent, which still seems pretty low.
This indicated that the distribution of "point winning rate" is heavily centered on 50%, which is something I would like to see (for non-math people, if you calculate the percentage of points won by all players, and just count how many won 50%, how many won 50,1%, so on and so forth...). In other words, tennis is utterly competitive. Which circles back to your original point: those guys that won so many matches (and so many important matches) might do something different in big points.
But I do think that there are different ways to get there, and the Fedal case is a great example. Federer, for me, creates margin by being "streaky", in the sense that in a few moments during a match he raises his level absurdly, than wins, say, 75% of the points, gets (and saves) the breaks, gets the lead and then sails from there. Nadal, on the other hand, is more about the pressure points (as is Djokovic). Well, that is my guess, anyway.
