I don't know the recordings you refer to and I can't hear your room but there are several things that COULD be going on. I say could because it would take careful measurement and listening to determine it. The most general reason is that those recordings with the most improvement have sensitive musical info at the frequencies that your room and/or speakers are most "off" combined with placement of those instruments on the left to right soundstage so that the effect of the room interactions are most noticeable. Also, the musical info is not buried in the mix. I will use an example from my situation with a Sigtech and Dunlavy V's. I have asymmetric bass response in several narrow bands starting at 80 and ending around 200hz due to a large Middle Eastern style arched door in the plaster wall. If I listen to a well-recorded string quartet with the cello in its normal position on the right of the sound stage the cello sounds thin and hazy. Reverse the channels and the quartet sound fine. Kick in the Sigtech and it sounds fine. Play a quartet recording with an extra wide soundstage or a tight soundstage and the effect almost goes away. Second possibility, you may have a general rise in response from the lower midrange (room or speaker). Recordings mixed with a similar rise in response (not uncommon with pop and rock) will sound bad. Recordings without the rise will sound OK. Both will sound better with the DSP turned on and correcting for that rise. These are the real effects I've heard and measured. Of course this may have nothing to do with your specific situation. and correcting for that rise.
One simple test is to play the recordings in question in mono or on headphones and note the difference.
One simple test is to play the recordings in question in mono or on headphones and note the difference.