xG is not a BS stat.. but interpreting it after a single game is pointless. I think xG tells a story after 15, 20, 30 games or so.
For example, if yoh toss a coin, the expected probability is 50:50 for heads or tails. If you guess tails and heads come up, is the coin rigged? Course not, you just need a bigger sample.
If we had a higher xG than Bournemouth and didn't win, unlike the overly simplistic example above, it does tell us something. But we should really be looking at xG, xA over a longer stretch of games.
If on a probabilistic basis, as in, if we create chances that shoud result in 50 goals, but we only score 35, then that's clearly a problem. Same for xA. I would hazard a guess we underperform our xG by a decent %, and also have a higher goals conceded than our xA. Now you probably don't need xG and xA to know we have too many individual mistakes at the back and we aren't ruthless up front, but there's certainly some value to take in drilling down to those stats.
Who are the main culprits? What flanks does our xA underperformance come from.. etc etc?