You have to go outside the stat set you commented on to support your argument with PER.
I was merely responding to your comment about Maya's stats, not relitigating the Starling value conversation. I think any hopes of Starling turning the corner have passed. We shall see, but nothing in the above statistics indicate he is the weakest link as you stated.
You choosing to bring in other statistical measures, like PER, is moving the goalposts.
Not every discussion has to be dragged kicking and screaming into a "Starling sucks" referendum.
But, since you brought it up, I agree, most of the advanced metrics think Starling stinks. By a lot of statistics, he is having his worst year of his career. But they all agree his defense is better than ever before, which to me is faint praise, but passes the eye test. Starling has bad defensive instincts, any time there is a 50-50 "should I go left or right" situation on defense, he goes the wrong way 80% of the time. It's uncanny. But he is working harder than ever before, and is battling through screens and staying in front of his man more than he has in the past. He also has help from Kyle, unlike the previous years at Syracuse. Having literally no center as a sophomore and Lampkin as a junior probably hurt Starling's DRtg substantially, because when he did get beat, he didn't have any help. This year he at least has Kyle and other shotblocking big men to help him out.
The thing I worry about with Starling is that I really thought having a true point guard would help him become more efficient, and all it has done is reduce his usage and minutes.