What? You want to compare the success of particular draft classes? What a radical thought that I in no way would've considered!!!!
That right there says I understand that there's value in comparing draft class performance, so the fact that you spent most of your post lecturing about the idea that comparing draft classes has value as if you're refuting something I said to the contrary is at best misguided and at worst a straw man.
The entirety of my point was to say that
comparing draft classes has value, but only in contexts where the comparison carries actual useful weight. Like for a career, or a multi-season stretch of time (like, say, performance through every player's first X full seasons in the league), or if you want to butt up against the abyss of "so narrow it's meaningless" without going over, then maybe you compare individual peak seasons. in hindsight.
The only time anyone talks about a single season in a vacuum with regards to a draft class would be for the season that represents the rookie year for a significant number of its notable members (which will require hindsight), or barring that then the year immediately after the draft to compare the players that make the leap right from the amateur ranks to the NHL. And the fact that I have to specify that it could be one of any number of possible seasons based on differing rates at which members of a class make it to the NHL is part of the reason that such measures are pointless: because they create far too much noise to muddy the value of the info observed.
The big reason for this is that every prospect matures at different rates, and one particular season can represent a variety of different growth points in the careers of the players involved.
Nobody's going to go back and say "who was the highest scoring member of the 1991 draft during the 1995-96 season?" (it doesn't matter that it was Peter Forsberg. Yes it means Peter Forsberg was great that year. But there are tons of other ways to establish that he was one of the best players from the 91 draft besides isolating one mid-career season against other players who will have had differing amounts of NHL experience to him by that point). You might want to know who was the leading
career scorer for that draft by the end of the 95-96 season. Or look at who had the best rookie season regardless of when in that span it took place. But asking which players from one particular class had the best year in one particular year offers you no useful or valuable info that you couldn't get through other more sensible sets of data.
To draw another baseball parallel, this feels a lot like the argument around the value of pitcher wins or RBIs as measures of player performance/ability. Can you potentially look at those stats and identify great players? Sure. Are there significant flaws that threaten the reliability or validity of those numbers? Yes, that's the big problem with them. And are there better measures you can use instead that don't have the same amount of noise, distortion, or randomness that create those uncertainties? Undoubtedly.