The reason we don't see a killer "business" app yet is because of the resolution. Once we solve this, and figure out a way to scan and accurately represent the location of interfaces like keyboards, we will see a rush of killer apps.
As for games... locomotion constraints really do a number on what works in practice. Carmack's inside-out tracking seems to be coming along well, but then the problem shifts to securing a large enough play area.
Yeah resolution is basically the problem here. My future of computing idea is essentially a desk with a headset, keyboard and mouse.
I'd buy and use this setup if it had the resolution for a lot of money and take it to work (noisy office, literally no reason for me to see around me, open offices have a lot to answer for).
As for games... locomotion constraints really do a number on what works in practice. Carmack's inside-out tracking seems to be coming along well, but then the problem shifts to securing a large enough play area.