Good job, stay safe!
I do think that having different font settings for the modes is confusing, at least it doesn't add value.
As a live test I imported over 1000 .cho files (several sets in several groups). If i were to render a subset, multiple select really becomes important. Also, for the case of rendering, it would be nice if the order of files could be reordered (shuffling entries). Think a setlist, where the order of songs matter.
When I click on [+] By Status, I get an expanded list:
Note that the sublists are expanded while stil displaying a
Maybe worth considering: add cpdb specific meta data (status, group) to the songs.
When extracting a song you can ask whether the custom extensions must be retained, so they are useful upon subsequent import.
In the META_DATA (in the zip) please include the filename as well. Even though not used by the archiver, it makes visual inspection easier :).
Using a comma to separate the fields and the groups blocks augmenting the entries in the META_DATA. Maybe it is better to use a distinct separator for the groups. See also the numerous discussions on the format of CSV files. Best is probably to switch to a well defined and supported and flexible format like JSON.
Happy thanksgiving (whatever it is)!