I can understand why reliable use of the advanced features might require specific formatting of the epub file, and given the wide range of variation used by different sources, the Kobo developers would currently caution against trying to use those features on any old epub. The results might well be hit and miss, and the misses might damage your database. But obviously this is the type of problem which is easily solved with more careful programming to detect and handle format variations, so let's wait and see if the Kobo team has what it takes.
