The current most popular theory seems is that they are using the usual VR tricks, plus have the ability to track the eye so that the eye's accommodation (focus) matches the distance that the object is supposed to be from. From what I remember with my experiments with stereoscopy this should reduce VR sickness in about 10-20% of the population for which accommodation mismatching with stereopsis causes issues.
Note that this effect is why 99% of the content in feature-length 3d movies appears at an apparent distance of more than 10 feet from you; the amount that accommodation falls off rapidly with distance, so it's a much less strong effect at that distance.
Note that this effect is why 99% of the content in feature-length 3d movies appears at an apparent distance of more than 10 feet from you; the amount that accommodation falls off rapidly with distance, so it's a much less strong effect at that distance.