No one is claiming that PDO is entirely attributable to luck, only that its extreme outliers are so improbable to sustain and repeat that you can attribute such a team's extreme success or failure, at least in big part, to luck.
Sure, the Canucks played well last year, but the production of many of their players was highly improbable compared to other players on other teams with similar ice time and that generated a similar amount of chances. This year that has corrected itself and the Canucks are mediocre despite having largely the same roster.
This year nobody expected the Caps to be this high up the standings and it's the same thing. Let them, like the Canucks, regress to middle-of-the-pack PDO and they will fall through the standings despite having similar underlying numbers.
Some people (like Canucks fans) might attribute this to gambler's fallacy, but that is not correct. No one is saying that because you were lucky one year or extended stretch of games, you will be unlucky the next. Only that you will regress toward the mean, which you will.
But the crux of the biscuit isn't just that it contains an extended run of fortuitous results; it is obviously that it is extremely frustrating to lose to such a team, especially since the fans of said team will typically rub it in and claim it's, in fact, not luck which comes across as a toxic cocktail of smirk and delusion.