AR products like the Apple Vision Pro already have a much stronger value proposition for casuals than immersive VR. A quality lens on a comfortable headset could replace every monitor you would ever have to buy, with no stands or cables, usable in practically any position.
It could unshackle the screen from the PC completely. You could keep your computer in any room and access it with wireless peripherals. Commercially available low-latency networks already support this at a usable level.
It's also a longjump in the direction of consumer-grade cloud-computing-as-a-service, which brands like Google and Microsoft have been trying to sell in specific niches (gaming, mainly) for years. Consumers dislike cloud computing in principle because it gives custody of their personal computer, with all their data, to a third party. They also dislike it because it is currently accessed inelegantly through another computer.
AR lenses obfuscate the first problem by hiding the computer out of sight, and completely solve the second problem by eliminating the need for a physical computer near the viewport(s) in a more versatile way than a similar-functioning standing monitor could provide.
So that's a weird take from Spencer. AR is the same level of virtual as regular computing, but much more comfortable. Strictly as or less "inherently dystopian" than what we already do every day. He said it about the "virtual realm" but called out Apple specifically so yeah if that's his angle, I don't think he thought about it like this.
TL;DR Apple Vision Pro is a hardware evolution with clear benefits to consumers and different but also clear benefits to cloud computing providers.