Great thread! Some comments:
Google Cardboard
I've tested a dozen or more and all of them are complete garbage except for one: the PowisVR. First off, most cardboard units have way too much curvature in the biconvex lenses and b. place those lenses way too close to the device. One or both is a recipe for a crap experience. The PowisVR has an appropriate curvature, and each lens is threaded to allow for adjustment. It's the only cardboard style unit that offers that feature.
GearVR
Pretty overrated. That's not to say it's junk, far from it. However, there is absolutely zero difference in the overall optic quality vs the PowisVR unit. They aren't using any better glass technology. Does the experience offer more immersion because of less light leaks? Yes, but not dramatically so... With GearVR and cardboard, you are still at the mercy of the resolution of your phone. In fact, if you are using a browser based experience, the browser will further reduce the resolution somewhat. There's no way that the pricetag for GearVR, not to mention the fact that you have to use Samsung phones, justifies our use of the product in widespread client engagements for fixed point rendering panos.
Also, and a lot of folks don't understand this, you are never going to do realtime VR with GearVR. It's fixed point rendering VR, only. The phones have a long way to go before they could handle any serious realtime work.
Oculus Rift
Great unit. The Vive is better.
HTC Vive
Not much to say there other than it's the way to go. There's a zillion reasons why, but I'll point out that countless hardware manufacturers and architecture firms have adopted the Vive platform over Oculus. Oculus = consumer market. Vive = seriousness. heh.
Hololens
Not ready for primetime by a longshot. The limited field of view is pretty horrible. While you can view any FBX file, only Unity is supported to create anything interactive. It's the future, for sure, but right now the promise outweighs the delivery. Maybe the next version...
Customer/Client Experience
There's absolutely nothing like giving branded PowisVR viewers away to the client after the presentation. The 'take it with you' or 'leave behind' approach is priceless. They love it. Their co-workers back at the office love it. Their kids and friends love it when they take it home. We've learned very, very quickly that the overall experience and...I'll say it... 'fun factor' cannot be underestimated. We tell some great stories with our virtual tours using KRPANO, but equally as important to those stories is the fun the clients have. Can you put a pricetag on that? No you cannot. hah.
I'll say this as well, because it's worth mentioning: Of the dozens of clients we've presented to with virtual tours, easily less than 1% of ever experienced VR or AR. This means that we are their first introduction, and that's the pretty awesome thing - especially when it goes well. 5 years from now everyone and their grandmother will have VR and AR but, for now, it's still magic to pretty much everyone.
Multiformat Approach
I tell folks a lot that it's important to keep in mind that the Rift and Vive aren't just for photorealistic real time experiences. Far, far from it. There's half a dozen ways to get a realtime experience going in both Revit and SketchUp. We need to set aside our innate need for photorealism and recognize that realtime VR as a presentation tool tells great design stories that don't need realism. Spacial relationships, content, density of space, etc etc are all fantastic to present with realtime VR. We will go from a beautifully rendered fixed point VR experience with KRPANO and cardboard viewers, to a Sketchup realtime VR tour during the exact same presentation. It doesn't have to be one or the other.
Game Engine Madness
We're adopting Stingray because the workflow from Max/Vray to Unity or Unreal is complete garbage for anyone that deals with serious real world deadlines. We don't have months to delicately pamper FBX files and Unity materials in order to show off yet another condo loft VR example. We have a matter of weeks to build full office floor experiences. We have a ways to go until we are comfortable enough to put these interactive VR presentations in front of a client, but it's very very clear to us that AutoDesk is serious about supporting the architecture and interiors world. Unity could care less about how to get something from Max/Vray to their platform. AutoDesk's integration and live link between Max and Stingray is huge. Purists from Unity and Unreal camps hate Stingray. Honestly, it doesn't look as good. Yet. While Unity and Unreal continue to ignore the crushing deadline demands of the realtime archviz world, AutoDesk is right out front providing tutorial after tutorial for great workflow ideas.
As I mentioned yesterday in a thread over at Chaos, at the end of the day - given our project schedule demands - I'm going to gravitate towards whatever company and platform wants to make my life easier.