Interesting topic.
It would be interesting to get a comparison between this method, using a built in renderer for SKP such as Vray or Thea and just plain old SKP viewport. What I mean, is how we might understand the performative nature of light, and also the sense of scale and immersion from these programs particularly with a view to outlay of time/effort.
I suspect UE4 would have the scale/immersion won hands down (inhabited camera vs detached from human experience), but lighting quality would almost certainly be taken out by a non “RT” traditional renderer (for obvious reasons) - unreal can create absolutely beautiful lighting conditions but they are rarely connected to physical, experienced reality, at least as much as a brute force GI engine can be.
I always envisioned UE4 being best used either very early on in the massing/early concept stage (minimal hassle of unwrapping, easy to iterate) or at the very end (when major design decisions have been made, truer sense of occupation, how a space is experienced when filled with the things of every day life). It doesn’t really suit a truly iterative design process due to the disconnect between the creation of virtual space (skp) and the experience of it (UE4) - it makes it harder to be really critical of the design work I would suspect.
Just my opinion of course - I spent a couple months looking into this for my masters, ended up discarding the idea of it as a design tool and sticking to a research through design process (art gallery). It’s a big topic, I reckon in coming years when there is a better flow between programs this will change.