Illusion is Part of the Definition
We created a scenario in immersive virtual reality, relying on research on body ownership, to give people the illusion that they were in a cell standing in a stress position, while hearing an interrogation going on in the cell next door. There is an article about it in El Pais: Presos de un Guantánamo virtual, and a youtube video: Towards Immersive Journalism: The IPSRESS Experience.
My new paper attempting to give my current understanding of 'presence' has been formally accepted for the Philosophical Transactions of the Royal Society, Series B. This is about deconstructing the concept of presence into two orthogonal concepts: 'place illusion' - the original idea of having a strong illusion of being in the virtual place, and 'plausibility', the illusion that what is happening is really happening. Of course these are illusions, no one really believes that they are in the virtual place or that what is happening is real - the fact that these are illusions is part of the very definition. Also the paper discusses what I think is a good way to understand 'immersion'. Altogether there are four concepts: Place Illusion (PI, based on sensorimotor contingencies), Plausibility (Psi, based ultimately on various correlations between actions and events), immersion (forming a simulation hierarchy) and the virtual body which is the intersection between PI and Psi. I think that it is relatively clear to me now, ideas that I've been trying to find for many years. 'Presence' in the sense of place illusion is not and never really was the main problem - it can be relatively easily achieved most of the time with the right display and especially tracking systems. The real problem is Plausibility. The pre-publication version of the paper can be downloaded from my publication website.