RJ,
The win shares stats are accumulated from this THIS SEASON - 25 or so games in. They aren't projections about a player's career or trying to predict future value. In case you haven't noticed, Dallas has struggled this season. They are 12-14. The Knicks are 11-14 (BTW, Mitchell Robinson led the NBA in FG% last year with a ridiculous 72% from the field: Robinson rebounds and blocks shots; he's a 7 footer, that can run and play defense; he is a valuable player and has played to his role this year; no one claims that he will be franchise altering player). Luka would admit that he hasn't played well and the Mavs have not played as expected this year; the stats reflect that. Last year, Luka led a late season surge for Dallas, maybe he will do the same this year, but right now -- 25 games into the season, Luka has not been a franchise changing player and that is reflected in the Mavs record and the win shares stat.
The win shares stat was not posted (and please check the language of all posts on this thread) to argue that Mitchell Robinson (or John Collins) is a franchise changing player or that they are more valuable future assets than Luka. Win shares is simply an analytic snap shot of how players have performed over a specified period of time (e.g., a season or two seasons or a career). Rather than attempting to understand that simple concept, you claim the stat is worthless, when it only reflects what is obvious: Luka has been up and down to start this season.