Originally Posted by robko
This is purely a (semi) educated guess but probably one of the fields is legacy field that was used in either e-reader firmware or one of the IOS/Android apps. Sometimes its probably easier to just leave it (kinda like an appendix) than remove it and risk possible unexpected weirdness -- which would include both of them needing to be identical suggesting that both of them do get used somewhere/somehow.
I agree that may be a reason why it is so, but then you would make one a calculated field set to be always equal to the other. A need to keep it hanging around just proves, once again, that kobo have some rather weak software development practices. If, say, the Touch only works properly when they are the same, then it should only read one of them, and use that value only. The very fact that it get's weird, as reported, means there is old code, within the Touch, reading the other one as well - and that shouldn't happen within the development on one platform.
btw: I am sure they recently found that a human appendix does have beneficial uses to you