The 3D of today is a gimmick. It's in a no mans land of 2D and 3D... lower frame rates then it needs to work well with movies and videos...
Slows down the frame rate on games.
Adds only a few minutes of 3D pop before your brain adjusts and forgets about it.
I mean, I get why some of you like it. I certainly don't mind it myself.
But when you contrast it to 3D end game... i.e. VR, you'll understand why I hold the current form of stereoscopic 3D is such low regard.
That said, I'm curious to see how they'll eventually adapt film and cinema to cope with the strengths of VR.
Will they have hand held high speed, high resolution camera array mounts used to capture subjects from a variety of angles with enough detail that it can be interpolated into a 3D model?
Even in such a setup, there are going to be areas that the cameras don't capture... so presumably, if you shift your head far enough, you're going to see artifacts that result from a lack of information.
To give you an idea of what I'm talking about... here's a link to a process called DTAM (Dense Tracking and Mapping), that uses a webcam to recreate the space in a stitched together point cloud 3D.
http://www.youtube.com/watch?v=Df9WhgibCQA
Presumably with 10-15 years of development, such a system could reasonably extrapolate details into the unseen areas - it won't be able to tell exactly what's there... but it'll be able to at least clean up occlusions and extrapolate textures and edges into their logical corners.
When combined with software that intelligently deals with an excess of movement (i.e. the scene is intended to be viewed from a fixed location; if you start moving around, the scene will follow you, rather than keep giving you information on what's in the environment), it should work quite well.
The desired effect should be akin to having the movie play out in front of you like a personal theatrical show, except keeping the significant advantages of theatrical movies - framing, pacing, special effects, etc will still be part of such a movie effect.