I have come across the following interesting behaviour in switching a large number of documents from being imported to being indexed, in the case where each document is in a different group.
The documents all have a common identifier in their names, so they can be selected collectively in the search window after searching on that identifier. They can then be dragged into a Finder folder and deleted from the DT search window, ready to be indexed back into the database from the Finder folder. But the documents were scattered across different groups originally, and need to be indexed back into their original groups, so what to do?
It turns out that if you simply index them into the Inbox of the database (through a cmd-ctrl drag and drop), then DT creates replicants of each of the indexed documents, one in the Inbox and one back in its original group! Simply delete the replicants in the Inbox and you are left with indexed documents in their original groups where you previously had the imported ones.
I am not sure if this behaviour is well known or not, but it is new to me (and I’m not sure how DT knows where the exported document originally came from in the database, once it is in the OS X file system and deleted from the database). I trust that it is intentional and not just a happy accident that will break in future!