I do think there's a bit of an argument to be made about the sustainability of it all but only time will tell.
Looking at something like that it's quite unusual for a team to score 4 from a combined 1.38 xG across 3 games and concede 0 from a 3.23.
Having said that I do put caveats in there. Top forwards score at a better rate than xG consistently. Son & Kane are top forwards. Top goalies save shots at a better than expected rate consistently. Lloris is handy. Still part of me looks at that and thinks they had the rub of the green a bit there.
Of course, scoring early in the 2 they won there is part of what went on to be. Neither were huge chances at all but seeing as they did score it arguably allowed them to play out the rest of the game the way they did, one in front, soaking up what the oppo can muster which in Arsenal's case was barely anything and not even necessarily looking to counter too hard, just when there was glaring opportunities to do so. They wouldn't necessarily have played that way should the early goal not have come in those games although they did against Chelsea so who really knows.
I do like xG but I think it is possible to break it (or for it to be wrong for certain tactical setups). Klopp kind of has the past couple of seasons, maybe Mourinho can too.
Swings and roundabouts too to an extent on xG, sometimes it goes with you, sometimes against. Probably merited at least a point against Everton on opening day looking at it but lost and should have beat Newcastle but drew. Southampton and Burnley are games they won but maybe should have drew on that measure.