I completely agree with Fons' point about wavelength and its
relationship to objects within the room. But, I've also seen
this mirror trick, "image method", in audio literature, for instance:
http://www.sgm-audio.com/research/rir/rir.html See note 
I think I also saw it in Philip Newell's studio design book.
To Fons' point, none of what I've ever seen on this topic ever
included room contents, and this has always been presented as an
_efficient_ approximation. For that efficiency some accuracy has
to be left out.
My sense is that this might have some usefulness, but not be the
final word on the subject.
Linux-audio-user mailing list