Vladimir, as I see it, plenoptic imaging is not 3D at all (it gives not a realistic spatial impression at a single glance), and holography for true color real size scenes it completely unrealistic. The obvious feasible and technically ideal (no compromises enforced) method for 3D television (also future) is the shutter-glasses method.
I think that Bahram Javidi and plentiful count of the other researchers would be disagree with your opinion that plenoptic imaging(integral imaging) is not 3D. I also think that japanese channel NHK, which want to implement this technique is also will not agree with you.
Well, maybe a very interesting debate is what we all understand as 3D...I find myself often discussing on that. Is reconstruction of multiple planes in 2D able to be called 3D? Or is it depth perception by human eyes... hours of debate may follow. However in the end it comes to: 3D data or 3D perception? And the trick is both need each other.
If I had to bet between your two options (I mean other techniques, e.g., time of flight or stereoscopy, are possibly not total outsiders here) I'd bet on holography, essentially because of the limits in spatial resolution there were in plenoptic imaging, at least when I analyzed the technique two years ago. Maybe things have evolved extraordinarily... But sure there is a lot of space for all techniques in different applications, as usually.