To me xG never truly reflects on the game in terms of how a team should of scored more. Like you can get loads of headers from corners and can easily pad those stats out. Again like I said in the other post you can have a low xG but can still score more than the opposition.
xG is not really a useful indicator in the context of a single game any more than possession, or number of shots is. And it's not supposed to reflect the scoreline as you suggest, it is a statistical measure of the quality and quantity of chances created. Which is a separate but useful measure.
I think people just totally misunderstand what it is and what it is used for. Likely the media's fault as much as anyone's. People don't have debates about how useful "Shots on Target" is as a measure, xG is just another such stat. It is an objective measure of one aspect of a team's performance.
The reason it is useful (and literally every team in league football uses it) is because statistics is all about sample size. Over one game a team can easily outperform or underperform their xG - but over a long enough time horizon teams will always revert to the mean, there is nowhere to hide from the law of averages. xG is designed specifically to be the average.
It's like how you can win a few games of roulette but if you play for long enough the house always wins.
xG is used as an early warning indicator for teams, telling them if their results are really justified on the basis of their performance or if they are papering over the cracks by getting lucky. It is provably one of the best statistics for this purpose.
Luckily for us our coaching team understands this - which is why our xG has outperformed every other team in the league every single season since Pep arrived. Our success isn't lucky, it's the law of averages. In our game of roulette we are the house.