Despite the subjectivity, the track record of correlation between expected goals and playoff participants and Cup winners is very strong (as far as expected goals from transparent data sources). There’s other factors, but nothing as strongly as that of which I am aware. Special teams dominance on both ends of the ice for example might show favoritism towards winning teams, but it hasn’t correlated with being a necessity. 5v5 play pretty much has until it shows out otherwise.
If there’s one source that deviates heavily from the others, I’d be leery of that source and how they are getting what they are getting. Might be bias or questionable methodology. It could also be they are innovative and ahead of the curve, but there’s nothing that suggests the other sources have been errant.