One reason why user-generated 3D content in general looks bad is that for a long time, "professional" content typically has lighting applied to surfaces as a te

One reason why user-generated 3D content in general looks bad is that for a long time, "professional" content typically has lighting applied to surfaces as a texture, and that texture is expensive to compute and requires complicated software. User-generated models on the other hand (like one might find in Second Life) don't look that great in comparison because the lighting is simplistic and none of the models have pre-baked lighting textures. As real-time global illumination becomes faster and better, I think we could get to a place where user-generated content looks pretty good by default, because it's always being re-calculated. I haven't really kept up with what the latest games are doing these days, so I don't know what the state of the art is, but there have been some pretty cool demos of real-time path-tracing. It may be awhile before that sort of thing is fast enough to meet the latency and resolution requirements of VR without being grainy, but I think we'll get there eventually.