I've been trying to stay out of this (and other similar) discussions but I have to through my 2 cents in.
Far be it for me to downgrade the importance of room accoustics (see my article
here), the statement:
Unless your room is an absolute abomination, it doesn't make a whole lot of difference.
is absolutely accurate.
Hold on, hold on.....hear me out.
There are two ways researchers deal with bias in experiments:
1) Reduce/eliminate it.
2) Control for it.
#1 is self explanatory. #2 means that the researcher knows that the bias is a problem in the experiment and makes every effort to minimize its effects.
So, for example, if children tested at the beginning of the day are known to score higher on a test as compared to children tested after lunch, a researcher would make sure to test all the students at the same time of day (either time, doesn't matter as long as all are tested at the same time).
In the exact same way, as long as the same room is used to test every reviewed pair of speakers, then the room effects will be the same for all. If the room accentuates the highs, it will accentuate the highs for all speaker set in the room, if it has a 60hz bump, then it will have the same bump for every reviewed pair, etc. Now, if you review a pair of speakers in a room, add room treatments, then review a second pair of speakers, this would be adding bias to the experiment instead of holding it constant for all.