On the website, the demo has the person drinking coffee, the youtube window is the wrong ratio, the text is too tall. For me, I really want picture perfect rendering before jumping. I don't want to look at badly rendered stuff, its part of the reason linux was so hard back in the day, getting your text to render and screen resolution to work right.
Now, for me, it looks like I can fit about as many terminals on "screen" as I can on two 1280x1024 screens. In all the demos, there are at most 5 windows open. On my current screen I have one browser, and 8 terminals, and there is still loads more space.
The thing that makes me a little sad is that we still all appear to be stuck at rendering everything on the inside of a sphere. If we are in VR, then we don't have to limit our selves to laying out windows on a single primitive. Where are the virtual shelves? where is the quick change, what about hotspots to bring groups of windows back into near field.
We have unlimited z depth, surely this is time to start using it? unlike a normal screen, we have parallax and gaze sensors, we really need to start using them.
(it looks like they don't have room tracking, so feels like they have limited 6dof source: coffee video, I'd expect much more sideways translation if they had proper headset tracking)