Honestly the pass reception is the key to separating the great passers from the good from the bad. Hockey players rarely flat out miss a pass to a player. They may give it to them too hard, or just behind them, or off the skate, etc, but they usually don't flat out miss them.
So if you get the pass reception to work properly you'd see the difference between passers. Crosby would be able to give players a pass right on their stick at the right tempo, whereas a lesser passer may put it too hard and out in front, causing the puck to skip away a few feet off the players stick, etc. Pass reception is the key.
The other issue is that vision plays a huge role in separating players, and that is a lot harder to simulate in a video game. Certain players simply see the ice better, they know where their teammate is going to be and they can anticipate the play and make a pass a lesser skilled player wouldn't even think to make. Or a less skilled player may make a bad decision that a good player rarely makes and try to force the puck somewhere they shouldn't.
For the user controlled team, this aspect is immediately negated since you control every player when they have the puck. And for the AI team, I imagine it is hard to program logic like that (we all know the difficulties in programming AI to begin with). So in that regard, a huge part of a players passing skill is already very hard, if not impossible, to differentiate in a video game.
Which really leaves pass reception as the most important factor in differentiating passers.