The problem is not on the RIS but on the importing of RIS to Endnote. I tried importing RIS on a different reference manager and there were no problems with Greek and German characters.
Maybe the RIS looks like nonsense when you open it with notepad, but there are just the HTML Entities (hex) of Unicode Characters, which they should make sense when Endnote turns them back to Unicode. In my opinion, the real question is why the “Text translation: UTF-8” does not work when I import RIS files.
I think that the order of first and last name is not really an issue as it could be changed in Edit-> Import Filters-> Edit X Filter - > “Author Parsing”.
I don’t think that merging or downloading from somewhere else could be practical solutions as I am using JSTOR on daily basis and for hundreds of references.
I would appreciate an alternative to the to the Unicode transformation.