Replying here rather than PM response ... best have more than one set of eyes on something like this.
Emma is correct ... all might be lost, but it's possible that individual courses were backed up in the 'normal' backup file area ... which is filedir ... by 'contenthash' named files.
So you have downloaded trashdir and temp. If the files have been moved to trashdir (believe the default is 4 days before cron moves unlinked files to trashdir). Doubt the temp directory would have anything useful in it, but don't delete what you have right now.
trashdir might look like the filedir directory ... nothing humanly recognizable there. So using a tool to inspect file types one would need to find if the contenthash named files contained in the subdirectories of trashdir are really obscured .mbz files.
Here's the response of 'file -b' command on Linux server IF the contenthash named file is *possibly* a backup (a .mbz file):
gzip compressed data, from Unix
Whatever file that hits on - name will look like: 257e169518aa86ab11a25e8d51728c73 ... you don't know the course name but copy that 257e169518aa86ab11a25e8d51728c73 file out to a 'save' directory some where else on the system and rename it to possiblecoursebackup#1.mbz
Do same for others, if found, only changing the possiblecoursebackup number to 2, etc.
Then restore them ... one thing however ... those backups might contain students and data ... so expect to clean up the restored course ... hey, at least ya have a course back so don't complain bout the work!
Sorry ... gotta do this ... <begin finger wag> always backup </end finger wag>
Best of luck.
'spirit of sharing', Ken