If you have the time, it is a piece of cake to empirically test the issue. Record a short SMF of arranger play, maybe something with the piano as lead (seems to be one of the voices that show up problems the quickest).

Now record it to your computer in stereo. Record it one more time using ONLY the mono out. Now sum to mono the stereo file IN THE COMPUTER. Compare the computer mono file to the arranger mono file... (you might also double check to make sure both files have exactly the same peak amplitude, and adjust in the computer if one is a hair louder).

Be great to have a definitive answer to the issue, and this might also give you leverage with Yamaha to get them to change whatever they are doing internally to make the mono such a poor comparison. If it is widely spread around that there REALLY is something wrong with their summing electronics, they might get around to fixing it to beat the bad press...

Worth a try, anyway