I would probably argue in turn that this kind of stat isn't really sustainable on a year-over-year basis. Over- or under-performance of xG (attacking or defensive) can happen in an odd year but Inter allowing 22 goals through an xG of 33 is just plain good fortune. Juventus had an xG against of 31.77 and allowed a total of 31 goals. Two years ago we had an xG against of 36.05 and allowed 42 goals. This year we're at 13.69 and have allowed 13. There's still the broader issue that our xG against is on pace to be much higher than a year ago (we're on pace for 47.2), but our correlation between the two is much more logical this season than it was a year ago.
(And having a great goalkeeper only goes so far as a mitigating factor. Just one quick, somewhat rushed example is Bayern from the last three seasons. Bayern's xG against since 2021-'22 has been: 38.61, 37.01, and 35.22. Their GA in those respective seasons has been 37, 38, and 45, so roughly +2 better, -1 worse, and -10 worse.)