I think it would help if the idea of luck as "some outside force controlling whether you do well or not" is suspended. For statistical purposes, luck is just the assertion that some players have skills that are simply not sustainable or repeatable.
Take a look at Brad Marchand and Jordan Eberle.
Brad Marchand is having a career year, and his current shooting percentage is 26%. No one is saying that he hasn't earned goals, that he SHOULDN'T be scoring that many, because he has and he's earned them all. The difference is that, to a certain point, you just can't expect him to continuously shoot that high. Whether it's a stretch of bad bounces, running into hot goalies, defensemen keying in on him more, or opposing coaches scouting the Bruins more heavily and finding out how to shut them down, he's simply not going to shoot at that level in perpetuity. Saying he's "been lucky" is not saying he's secretly a bad player, it's just saying that he's currently demonstrating a skill which has been proven in the past to be unsustainable (i.e. not repeatable).
The "shot quality patrol" was in full force on the main board last year defending Jordan Eberle's ridiculous shooting percentage. They said he wasn't going to come back down to earth because he simply doesn't take bad quality shots. Now he's on pace for 20 goals and 58 points in an 82 game season.