So he made L2 worse to not improve L1. Seems like a legitimate gripe to me.
....though I suppose you could get really granular with it and look at what the lines were doing at the time he made the switch, but I hope for their own sake that nobody here has the time or patience to do that.
Broken down a little.
Second half of the season he played Rust and Crosby 40 more minutes than he did Rakell and Crosby.
Rakell and Crosby had a 54% goal share, Rust and Crosby had a 70% goal share.
Which means Rakell and Malkin play more than Rust and Malkin, and Rakell and Malkin have a 40% goal share and Rust and Malkin 58%.
That works out as maybe he should have stuck with Rakell with Sid and Rust with Malkin more, as that gets you +5, while the other way around only gets you +3. But I get him sticking the stronger player at the time with Sid, and not overly disturbing the pairing getting big results.
And I am somewhat shrug energy over whether he goes with the top 6 usage that gets +5 vs +3. That doesn't kill you.
Breaking it down more granularly could produce another picture again.
Point in case - to look at the end - Malkin-Rust don't have a game with over 6 minutes after 20th March. In the preceding 8 games where they get some 6 or more minutes together, they only have one goal. If it's not happening, why push it? Malkin-Rakell score in that 20th March game and don't go more than 3 games without a goal for the rest of the season. Ditto Crosby-Rust.
I can't speak for the whole of it as I am indeed out of patience but it looks quite reasonable in terms of timing and going with the hot hand.