Some of it is luck yeah, some of it can be patterns and actual team or player traits. As I mentioned with Kane for example, you look at his goals scored compared to xG over the course of a season (which generally over that time frame a striker is not lucky for the whole thing) and see he is markedly overperforming - its somewhat of a measure of how clinical a finisher he is and is able to score goals more often than he should be.
As for whether quantifying it is worthwhile - I guess that's personal opinion to some degree, but to me its definitely yes, and I suspect the problem most people have is lack of understanding, not it being a number.
When discussing who won a game, you could say team A scored more than team B. Or you could say the score was 4-1.
You could say team A took more shots than team B, or say it was 11 shots to 2.
Putting a value to it is way more informative (as long as the stat is relevant and accurately determined) than dealing in the binary of yes/no.
I agree that any coach relying on stats alone is asking for trouble. But using them in conjunction with your eyes is useful, and can help to highlight things you may wish to examine more closely.
I think too many people look at stats and think "Its either got to be a 100% complete and accurate reflection of the game or its useless" when really most of them are useful if you just regard them as a different way to look at it.