Not to throw paper into the fire .. but after all of this placement issue .. you also need to know what you might expect from these .. so, one simple starting point would be to plot the room. To do this you need math: f =1130/(d)x2 .. so frequency = 1130(speed of sound)/(distance)x2 .. frequency = 1130/(21')x2 ... 1130/42=@27hz. This would be the 1st room mode for say a 21' length, the 2nd for that 21' length = 27+27=54hz, the 3rd = 54+54= 108hz, the 4th = 108+108=216hz. Do this up to say the 12th set of calc's.
Now, do the same for the width of the room, and also the height of the room, take these three and plot a curve to see what frequency(s) are close to or the same .. these would be the standing waves you will have problems with .. the less the better, and the lower the initial starting room mode number for each dimension, the better. If you did this for say an 8x8x8 room, you'd quickly see why a square room is a problem.
All this being said, also keep in mind that the numbers will not be exact due to whatever construction was used in the room and whatever else is in the room .. however, it will afford a fair idea of what would be an easier room to record and/or mix in.
Please correct my math guys if I'm off in the numbers above, but I believe it's correct, and certainly the principle is valid, you agree??