I liked the immersive aspect, especially when I was able to view my controls or use different kinds of actuation. I also liked when I was able to see parts of myself embodied (such as my hands).
I was hyper-aware that I was in a created space. This was limiting in some way. It was like a copy of reality, which felt like reality but lacked many of its affordances. The ability to navigate the space was limited by the controls, and how the creators had designed the space. It was a little surreal and created a bit of dissonance, like my experience/ability to observe was limited by the creators of the space.
One way to exploit the embodied aspect is to add additional position sensors to your face, hands, and body, and have those map to embodied selves in the 3D context. The environment can have different opportunities to reflect those embodied selves – the view of your hands, reflection in a pool, appearance in a mirror, people’s reactions to you. I think Kafka’s The Metamorphosis and movies like District 9 would be very visceral if individuals could play those characters.
To overcome the sense of being in a created environment, virtual spaces could be constructed either entirely or relying heavily on direct mappings from physical inputs, etc. 3D cameras (though mapping these inputs into vector or 3-D objects might not exist or be advanced enough). Like navigating different real-world environments – in the deep ocean, open grasslands, different architectural landmarks etc.