I noticed that you cannot copy a folder that contains links (unsync’d files) and have it populate those files on the existing cloud side. While it works in explorer (new folder is created) only the newly created folder is sent up to the cloud and the links which are copied over neither can download a copy of the content they were previously pointed to nor do they cause a copy on the backend (cloud side) for the data to be duplicated.
I am guessing this is just a limitation. But it means you can only duplicate data when it exists locally the the oDrive installed machine.
Example
Amazon Cloud Folder contains
\Backups
\Backups\PC1.cloudf
Copying “Backups” to "Copy of Backups"
Yields a folder that has a “PC1.cloudf” but clicking the file gives an error that it can not be downloaded.
Additionally reviewing Amazon Cloud Drive on the web only the folder “Copy of Backups” is sync’d there is no “PC1” duplicated there.
This would lead me to believe that you must be aware of any “Unsyncs” you have done before copying any folders as you could potentially lose data being unaware of sub-tree unsync’d data.