I am huge skeptic of football stats for individuals, but do see value when put in broader context. Having said that, I have read a lot about xpg (and the many variants) and I actually think it is a very good measure for a TEAM’s performance relative to their opponent… you have to still put it in context, like anything else. For example, does a team tend to underperform xpg (ie they are sh*t at finishing - I’m looking at you Danny boy)… there is also the potential to make the TEAM xpg for a game customized to the individual performers (depending on stat sample sizes this is more/less valid) - I don’t think they actually do that, but it would be a way to normalize for each individual’s xpg.
Anyway, long-story short, I think xpg and delta xpg (I just made that up, but I suspect exists) are very reasonable ways to assess a TEAM’s performance over a game or a series or a season. I bet it tends to correlate extremely well with league position.