Why would anyone want to work with a helmet on their head at all times like this? It's like saying the endgame for computers is all communications will be over video phones like on the jetsons. It sounds like a cool scifi concept until you realize even if the technology is there people would often rather not be seen on camera when they talk.
First, note I said VR/AR, so not everyone needs a headset. Many workloads will work in AR, though the black pixel problem means these will be much harder to crack and so will probably arrive later.
But I think once the VR headsets get miniaturized a bit more (give it a generation or three) people will laugh when they think about the current models, just like we do about cell phones vs. the first satellite phones that were bigger than your head. At some point these will be as light as a pair of plastic sunglasses or goggles.
There is nothing forcing you to use VR for something like a call, where you don’t currently need a monitor. But I think we’ll see a tipping point where the face tracking gets across the canny valley and people stop saying “you need to meet someone in person to really connect”. At that point VR calls substitute for in-person meetings, not VCs on a screen.
Consider the move to remote work; if we can get a virtual meeting room to feel like whiteboarding in person, including gaze and expression detection, then you could bounce between meeting room with your distributed team and perfect immersive dev setup without leaving your seat.
For the median worker using a monitor I think the requirements to beat monitors are just good enough resolution for spreadsheets/email (we may be there next gen?), comfort (currently the crux), and a decent story on input passthrough (your physical keyboard rendered in VR? Something else? Seems tractable, we just haven’t standardized any options.)
I see three assumptions in your paragraph that need to be true in order for the technology to work. I'm ordering them by how likely I think they are to come true in the next decade.
* Headsets will have good enough resolution and be generally comfortable enough to replace monitors for office space.
* There is a solution to the pass through input problem that is acceptable for the average office worker. I don't think there is a solution of the passthrough problem that beats a keyboard/mouse, or even a laptop in a cafe. I think it's more likely the average office worker will accept a worse form of text input given the right conditions.
* It's possible to project a 3d image of myself while wearing a headset that doesn't include the headset and passes the uncanny valley. The uncanny valley is wide, and even AAA video games haven't cleared it yet.
That seems like a good list. For 1, I expect to update substantially (either for or against) after seeing how much of a jump Apple’s headset is. I view this one as inevitable unless something crazy like a complete end to progress on SoC density.
For 2, there are demos already; Immersed (and maybe Meta natively?) has a mode where it recognizes your keyboard (like 2 specific models, prototype) and positions the keyboard in VR. Not good enough for hunt and peck but if you touchtype this works. The Quest 3 seems to have better passthrough so again, this generation will provide a good steer. This one seems pretty easy though.
For 3, if we can do deepfakes we can do 3d photorealistic avatars. Can’t be more than 5-10 years away to render a face in real-time. Unreal already has some crazy tech with MetaHuman that would work here already I suspect, given enough compute.
VR headsets are just more compact than laptops, let alone desktops. If you could just replace workstations with a headset, though it will not likely happen in near terms, that’ll be nice.
