There's plenty of stuff I'd love to tinker with that's just way too expensive and/or take too much time to acquire. For instance, there are some 3D constructions I'd play with, but I don't have a 3D printer (it's a PITA to operate, and I gave pieces of my last one to a hackerspace anyway). I have some ideas for a hardware product I'd like to test out, but no SMD soldering skills, no space for it, none of the more expensive electronics debugging tools, and one or two components in the initial design that are expensive and would take a month to ship. I could afford it, but I can't justify it. I couldn't afford it when I first came up with an idea. I can't imagine your average teenage tinkerer affording it either.
Now imagine that in a half-decent VR environment with real physics, a circuit designer and a basic circuit simulator, I could build a simulacrum of my project in hours, and then iterate on it even faster. I could build and throw away 100s of prototypes, for free and without waiting for parts. In the end, I'd have some idea whether the project is worth making real, and I'd have circuit designs and firmware to test out in the real world.
Speaking of environments, I've been in love with Bret Victor's "Seeing Spaces" idea for a while[0]. I thought long and hard about what it would take to build one in our hackerspace, and realized that this idea needs a lot of prototyping beforehand. Given a set of measurements on the thing you're doing (and the imprecision of the way they're collected), will it be helpful? Will the way it's displayed be annoying? Does it let you explore the project space faster? What kind of affordances can be provided? It dawned to me that it doesn't make sense to start shopping for actual hardware to test it all out - VR could be used to quickly test concepts and gain a good idea of which solutions would work instead.
Now I'm slowly saving up for VR hardware that will work well with my (rather strong) correction glasses...
--