Speaking from anecdotal experience with online communities, when idling or toying at home during their free time, people actually engage socially with friends through chats, networks, games — pick your entertainment drug. They're everything but alone, and it's genuinely real if limited cognitively compared to a real physical interaction.
If you add voice chat to that, at the end of the day you really feel like you spent X hours with friend(s), there's just no other way to put it except for the physical location.
It's almost a given to me that with appropriate videoconf wall-sized (real-world size) screens in all living rooms, we'd just "invite" friends this way most of the time unless the physical interaction was necessary. Less maintenance, less friction to just spend a cool quality hour.
The 'effort' required to move out of your house is getting increasingly higher, your 'local maximum' would essentially converge to an actual maximum in a perfect virtual reality (of which the "80%" / good enough might be just around the next decade).
This is why you see people holding conversations in real life while still sneaking looks (or not sneaking) at their phones. They're junkies.
Speculating I'd guess that the low lift and reduced anxiety means more people get stuck in it as a local maximum even if they'd be happier otherwise.