As the supporting logic, usually frame buffer based, for "pseudo 3D" displays gradually improves it seems likely that demand will increase, for "real 3D" displays implemented in a reasonably cost effective manner.

In contrast to a 2D display the frame buffer of a 3D device would be much larger for similar resolution and colour discrimination, with corresponding implications for detailed rendering.

However with Terahertz processor nodes, possibly with thousands of cpus on the horizon this may not be an issue by the time practical real 3D devices emerge from the "lab".