Make DOWNLOAD_ASSEMBLIES store data more effeciently

[zachary-foster]

Currently, the DOWNLOAD_ASSEMBLIES function downloads a zipped archive, unzips it, and renames some files in the unzipped archive that are used for the process' output. Both the zipped and unzipped versions are retained, so this uses more than twice the storage it could use. We only need to store some of the files in the archive and those can be compressed again once they have been removed from the unzipped archive. All the other files can be deleted by the process.

Things to do:

• Instead of renaming files in place, move them to the root of the process work directory while renaming them
• Compress all the required files that were moved from the zipped archive
• Delete the original zipped archive and the unzipped version