Nobody outperforms or underperforms the xG forever, that's the benefit of it. You can lose a match and be unlucky. You can lose 3 matches and be unlucky, but eventually, those balance out given long enough.
It is like playing roulette in a casino, the house might lose on 5 spins - but there is success if they don't panic and change the rules when stats say eventually they win.
It gives a better understanding of longer-term patterns because we are all wildly biased when we look at matches. This is the fundamental principle behind Brentford and Brighton, that we all have wild bias and that stats help you to unpick them.
For example, it is no surprise that Brighton have a dearth of Ecuadorian talent - stats showed it was an underpenetrated scouting pool, producing higher quality prospects than people would expect. They go in and find cheap prospects before anybody else, Caicedo, Sarmiento and Estupinan.
If you look at Brentford and the way they play football, they rarely take shots unless they are in a high xG location, and they consistently look to play the ball into those locations. Thomas Frank takes substitution advice from a team of analysts who show the underlying analytics and make suggestions based on them.
Sent from my SM-S911B using Tapatalk