There is no such research based on controlled (!) listening tests I'm aware of, just a myriad of audiophoolish "facts" ("It's true because I heard it!") filling up tens of thousands (millions?) of posts on social media.
I believe that this is a difficult topic to explore in a truly controlled fashion. Moving a speaker closer to the front wall changes its frequency balance, tending to increasingly tilt up below 500-1000 Hz with proximity, as well as changing the comb filter with the first null increasing in frequency with proximity, so multiple variables come into play. If one does this in a room with stereo loudspeakers, then the listening position should be similarly moved closer to the front wall in order to maintain the subtended angle with respect to the loudspeakers, but then this may change the effects of room modes as the listening position changes with respect to nodes and antinodes. One could perhaps do this in an anechoic chamber by placing a rigid barrier behind the loudspeaker, though that doesn't change the balance and comb filter issues. I posted before a report (
https://augene.blogspot.com/2020/10/listening-in-anechoic-chamber.html) where the speakers were placed closer to the absorptive boundaries themselves, which could potentially address the balance and comb filter issues to some extent (as anechoic chambers are only rated down to a relative cut-off frequency), though no frequency response curves were provided.
However, sophisticated auralization setups like those referenced here (
https://audiomediainternational.com/bo-taps-genelec-for-new-virtual-reality-lab/) could allow for rapid comparison of different setups (for example, signals recorded in an anechoic chamber with a rigid boundary or else with loudspeaker proximity to an absorptive boundary). The Archimedes project did not investigate effects of front wall reflections, and I can't tell if it properly addressed the differences in frequency response with respect to the varying frequency spectrum of the reflections relative to direction from the on-axis.
What i mean with dimensionality is the width and depth and the positioning of sounds in that 'space' from the loudspeaker plane to (sometimes far) behind that plane. What to my ears occurs is that f.i. when putting speakers on each side of a big furniture the sound the depth collapses to the loudspeaker plane and the positioning in the width gets disturbed. So , in my view, it looks like the spatial cues in the music get distorted.
Putting speakers next to furniture introduces the possibility of diffraction effects. In this post (
https://www.audiosciencereview.com/...can-this-be-settled.29550/page-5#post-1057682), I referenced several sources who probably wouldn't be considered by most to fall under the "audiophool" category, namely Sigfried Linkwitz, Floyd Toole, and Geoff Martin. However, you would likely consider my post to be just more background noise.