Yutaka says that CJK compatibility character replacement is not planned for the near future, although he does understand that it would be useful for the Korean language.
I did revise your macro to do this, and without freezing for multiple documents, hopefully.
Your macro works fine, but I did make some changes to make it easier to maintain, if you need to modify it later on. In Uni_CJK_2.jsee, I create a list for the conversion pairs, and loop though it. Unicode characters can be written like “\u0000” if that is useful.
To fix the issue you had with the macro not able to process multiple files, Uni_CJK_3.jsee fixes this by opening one document at a time. It saves and closes the document when the script is finished with the document. I think it was a memory issue, so if that was the case, this should work.