It seems like a good idea to have a consistent naming approach.
Re ID3 tags, does the stats program handle diacritic marks? Obviously Dvořák doesn't, as a string, = Dvorak unless you do some processing.
Don't I know it. Diacriticals are a royal pain in the butt when doing scripting or C programming on Linux, especially as the ID3LIB library is a very old one that does not support this kind of thing.
means you have potential problems with the Bach family, not to mention Scarlatti(s), Clara Schumann, etc.
Tell me about it. Can't remember the number of times I've had to change my scripts to cope with double names, names with spaces, etc. It's all these exceptions and duplicate names that make it such a bitch to get right. The plain basics are covered easily enough, which we now did.