Google+ failed because Google didn't know what they wanted it to be. I don't think that's down to engineering. That's a product and marketing problem.
AR is a little different because it's not a product; Facebook's AR apps might fail, but Zuck's vision could still succeed if Facebook create (or acquire) tools for creating an AR ecosystem.
Google doesn't actually have many customer-facing product wins on the strength of their GUIs / UX to its credit, period.
Google search was minimalist. Android followed everyone else in the market, and hasn't iterated core UX much since. G Suite office apps were acquired, and mostly cloned Office online. Same with Gmail.
Maps is about the only heavily-used app, and I wouldn't call its design revolutionary (or even good).
Chrome is about the best example I can come up with, and it's mostly iterated and improved on the back end.
Maybe I'm missing something major? But it's not a market they compete in or are particularly good at.
(This isn't to say that the initial lift of desktop apps to the web wasn't a technical tour du force; just that blue sky GUIs have never been something they're known for)
Mass media social networks are a winner-take-all game. All but one candidates will fail, and the success of that winner is all but certain to be based on a mix of ideosyncratic or non-apparent characteristics, though founding cohort can be and has been a major success factor.
Two of the largest online social networks to date have emerged from the highly-selective university space: Usenet (UCLA, Stanford, Berkeley, MIT, UI, Rutgers, etc.), and Facebook (once: literally Harvard).
Google+ might have been able to grow from its tech roots. My sense is that Google's marketing-and-advertising focus (and community), as well as gross mismanagement, doomed its attempt, nudged with some active antipropaganda from other sources, notably Facebook. That's not to say it would have succeeded, but there were numerous self- and externally-imposed injuries.
