Otherwise I'm with you. Positional tracking was the thing that really made it awesome for me. There are so many tiny head movements that you make when you're looking around; if positional tracking isn't pretty much 100%, then it just feels wrong.
If you're targeting a game or other application for VR (and assuming generally similar hardware to DK2 or whatever), there are a lot of design decisions you can make to improve the experience. Something as simple as using larger text goes a long way. Color choices help, as does less focus on lists and menus in favor of gaze-based control.
Then there's the general stuff like "not trying to shoehorn that awesome first-person-shooter style that was so great on 2D screens into VR". Yeah, it's neat to look around in a FPS-style game but whereas seated or less frenetic experiences are gorgeous and immersive, asking someone to run around in first person while actually sitting in a chair is nauseating. I've seen all the omni-treadmills and other workarounds and I appreciate the way people are trying to engineer solutions but I feel like it misses the point. Some things just work well in VR while other things require a lot of tradeoffs or addons.
It's natural pixel style makes you not notice the resolution at all, and all of the text is large, and easy to read. The cube nature also really makes me aware of the 3D, which is nice.
It doesn't do anything to deal with the FPS/nausea problem, though I don't react very poorly to playing an FPS in the rift as long as it uses a keyhole style control with the mouse (half-life 2 is a prime example of that).
One other thing I like about minecraft on the rift is that it really shows off the true blacks you can get with the OLED display. It smears a bit, but it's still really rad to look into a mine shaft and have it fade to actual darkness (not sure if the consumer version is planning for OLED off hand).
I definitely agree that it'll be cool when game shops start really thinking about building their games for VR, and that they can work around VR weaknesses.
Weird scaling is another issue, when something was scaled to look right from 2D on a screen, but drops the ball in VR. I played Skyrim using Vireo and the mountains ended up looking like glorified hills, heh.
Another funny issue I had was playing Alien Isolation. VR made it super obvious that looking out through a spaceship window into "space" was actually looking at a black wall about twenty ft away.
Looking forward to wherever VR goes, either way.
That said, I think Crescent Bay would help to solve the problem.
It seems like there are still plenty of challenges to solve with VR devices. John Carmack recently spoke [1] at the University of Texas at Dallas and mentioned that designing for VR is all about tradeoffs--hardware just isn't powerful enough to reach the ideal combination of resolution, framerate, etc.
He also talked about how to design media to work well with the current generation of VR devices. For example he said to avoid accelerating the user, because this commonly causes users to feel sick (due to mismatch between what your eyes and inner ear are telling your brain). Moving at a constant velocity is fine.
The DK2 is pretty close, but if you use it a lot the tracking isn't good enough as are the screen resolution and the refresh rate.
All these problems have been addressed so it should be great!
If you've been playing with the DK1 or DK2, the consumer version is fundamentally better. Do not assume you know what the state of the art is until you've gotten your hands on a preview. I don't know exactly what the resolution differences are, and I don't care - it's not the relevant spec to compare. It's gotten to the point where resolution is "good enough" to trick the mind and anyone who says otherwise is bikeshedding. They've gotten the screen door effect under control. With head tracking implemented and drift solved as a side effect, two huge problems are gone. Their low-latency head tracking stops people from feeling nauseated like they did with DK1. I could barely stand without wanting to vomit after playing the Doom demo. The only game I could play for any length of time was a simple platform jumping game called Rift Rush.
I don't know what they consider their largest hurdles, but from my perspective the hardware/software problems are solved for now and they need excellent made-for-VR content. When done well, the effect is stunning - in one of that calmer, more exploration-oriented demos I just stared at the light glinting off a metal bowl as I moved around it for at least 30 seconds.
But the difference between "good" and "mind blowing" was often the result of things that nobody has thought about before. In the movie-like demos, sometimes I missed things because I had my back to the action. In VR movies, the user controls the camera and needs to be told where to look; that's a problem nobody making movies has ever faced before. Even Pixar's demo suffered from this - the action was too spread out for the screen's field of view.