Thank you for the quick response. I am no expert, but maybe this mode could help to some extent with "speaker simulation"? https://developer.android.com/reference/android/media/audiofx/Virtualizer#VIRTUALIZATION_MODE_TRANSAURAL
It is mentioned that it's primarily to be used with speakers, but I couldn't find what it does when used with headphones. Probably even the standard virtualization at the minimum level would help to mitigate the strict left-right separation in some songs.
Btw. I'm sorry if it's been mentioned already, but how could I achieve sound processing system wide? I can see for example YouTube Music and Poweramp in "Known players", but I can't add Youtube/Youtube Vanced here (I also tried the trick with minimizing and playing the video track from the notification). With other Equalizer apps I am able to use "global" mode with Youtube. It is usually with a slightly worse quality than when connected directly to the audio session, but at least I can use sound processing. Did I miss something in the settings?
Not sure if it matters, but phone is Xiaomi Mi 10T lite, MIUI 12, rooted stock Android 11 ROM.