Yes – this is an important topic. We talk about this a lot here.
There are certainly some approachability wins to using VR to learn how to layout levels and build worlds (compared to traditional CAD-style controls), but that is very much an auxiliary benefit to us. We’re aren’t doing this to make it easier for newcomers, we want to help VR developers build games more efficiently, while also exploring new methods of interactions with game worlds.
I don’t think that anyone really expects developers to suit up for hours at a time while editing content on current VR hardware. That is more of a long term possibility, and this is very a forward-looking project. The idea is to get in here on the ground floor and discover what works with the community. However we already can see with our own VR developers that being able to interact with the editor’s world while in VR (rather than just previewing the game) is really quite useful, even when limited to inspecting and tweaking for minutes at time.
Creating an entirely separate editing environment to allow users to tweak levels in VR might not be a good long term offering. Everything gets implemented twice. Once for desktop users, then again (presumably stripped down and streamlined) for VR editing. We don’t really want two separate entire user interfaces, and I don’t think we can effectively maintain two without compromising one or the other. Instantly switching back and forth between the regular editor and UE4’s editor in VR is really nice too!
There will of course be some VR-specific features we’ll need, but overall I’m hoping that exposing existing editor functionality to VR gives developers the most flexibility. Plus I’m very much expecting improvements to the regular editor user interface as a side effect of our VR work. The UI will become cleaner to make it easier to interact with in VR, new in-viewport gizmos will be made available to regular editing modes, and we’ve already been improving features like WidgetComponent that will benefit even non-VR developers. It’s clear that a singular product is at least worth a shot.