Sofas can have effects as absorbers/reflectors. Shelves, too, if they are full of solid objects like books, are effective reflectors (they are unlikely to be diffusers at sub-Schroeder frequencies).
Basically, the more massive the piece of furniture, the more likely it is to have some effect on sub-Schroeder response.
However, I don't think anyone is suggesting completely emptying the room of furniture before measuring, but rather just moving anything that's within a couple of feet of the microphone out of its way, because reflections from very near objects will have a serious impact on what the mic measures, while being unlikely to have as great a subjective impact on what a listener perceives (for example, because they are used to sitting on a couch and the reflections off the cushion behind them form a normal part of the soundfield, which should therefore not be compensated for).