The biggest problem with pure statistical lists like this is that it's putting every player under the same restrictions with no context. Kadri is a really good player, and despite the mocking of the thread, he's producing a solid amount given he's now 3rd on the depth chart.
The thing is, he had a really great, peak season in '17, where he held his won against top competition with mediocre wings, and produced like a top 20 center. He was legitimately a top 20 center in that season. Then last year, he started strong in the same position, then slumped, before finishing well after a PPG Marner was put on his wing. His goals held, but his points fell in a higher scoring year, and he didn't do as well against his competition. He was a borderline top line center last year.
So, Kadri has a career season at age 26 after never scoring more than 50, and follows it up with his 2nd best season at age 27. Pretty standard seasons to play your best hockey, and over those two seasons combined, given injuries and down years to other players, there's probably a case for him around 20th or so in terms of overall performance.
The problem with this though is that there's no context for how he should be expected to perform going forward. Given he was better in '17 than '18, and given that his peak year came at an age for a standard peak year, it seems reasonable to assume that he wasn't going to keep performing at that level, and that it was likely a spike in his career. Meanwhile, for other more established players, you can see how a down year in a small sample would make them look worse than they are, and over a larger sample you can spot the outliers. While players do improve and regress, and this is often the reason for cutting a sample short, I don't think any one formula is going to put every outlier, good or bad in context. Sometimes it's obvious a player is in decline or has reached a new level, and sometimes it's obvious there were other reasons for it, like linemates, or a coach or a lingering injury, or unsustainable underlying numbers, etc.
While these types of things are subjective, I think they're also necessary if you want to predict how someone is going to play in the near future, rather than just analyze what they did in the past. For someone like Kadri, the reasonable assumption should have been that '17 was likely a career year, and, given his play in '18, he would probably play more like a borderline 1st line center/high end 2nd line center the next couple years.