I actually never used +/- against Kanter. I used RPM, because it's designed to eliminate the variables that aren't controlled by the player.
As far as PER goes, it shows what I already know - Kanter's an excellent offensive player, and crashes the boards very well. What you have to realize is the knock on Kanter - the reason RPM shows him as a net negative player, the reason the Jazz instantly and significantly improved upon trading him, and the reason Billy Donovan is only playing him 20 minutes a game - all has nothing to do with anything that PER would reflect. PER doesn't measure defense. It doesn't show that what Kanter gets on one end of the court is being completely offset by the negative impact he has on a team's defense.
As far as PER goes, it shows what I already know - Kanter's an excellent offensive player, and crashes the boards very well. What you have to realize is the knock on Kanter - the reason RPM shows him as a net negative player, the reason the Jazz instantly and significantly improved upon trading him, and the reason Billy Donovan is only playing him 20 minutes a game - all has nothing to do with anything that PER would reflect. PER doesn't measure defense. It doesn't show that what Kanter gets on one end of the court is being completely offset by the negative impact he has on a team's defense.