Hi @ben. Sorry about the late response.
I really appreciate your response and as a developer myself I can completely understand the design choice made for the re-scan algorithm, creating the unique hash for each track and as a final step comparing and removing duplicates of that hash. Sounds reasonable and well thought out.
As you’ve mention, this does explain the seemingly weird count during update but does not explain the issue itself (i.e. why the same MP3 tracks are “invisible” when mixed in the same directory structure as other files, but show up when I move them to another location and add that as a new Folder).
Nonetheless, I now consider this a minor issue, due to the work-around I mentioned (placing the files outside the main library Folder). And I truly understand if you – and the dev team – focus your effort on more important fixes and adding new features (e.g. this would be a really nice one, in my opinion 
).
Thank you!