I noticed a couple of problems when work with epub3.
1. Plugin removes xml header from .opf file and some other files with no reason
This one: <?xml version='1.0' encoding='utf-8'?>
2. Remove non-dc metadata elements is my favorite setting, but in epub3 we got some reserved properties (
epub3 spec) which is not just personal garbage but related to book itself. I'm talking about meta for store Series name and index:
belongs-to-collection (for Series Name)
group-position (for Series Index)
collection-type (type Series or Set)
And Calibre use these three fields when convert to epub3 to store Series name and index.