I have also heard Colin’s work and it is magnificent. So I wouldn’t pretend to question his methods.

But to clarify, assuming you need 8 panel memories for each song, it seems to me that you are limited to 8 variations (24 if you switch banks) of voicing and rhythms and would have to select from those same 8 variations for all 20 songs. And, that you have identical panel memories in each of the 20 SD card locations, each of which has a different song name. Since you are using SD memory locations for each of the 20 songs, couldn’t you also select unique voicing and rhythm variations for each of the songs and put that in the 20 memory locations? You would then have unique setups for each song. Maybe I do not fully understand what you wrote. It doesn’t seem to me that it matters whether you are playing by ear or reading the dots.