Bottom line
Deja Dup indeed is intended for you to be able to restore all data if the original data accidentally gets lost.
How does it work
Tar archives preserve all data, file names and attributes
Deja Dup uses duplicity as its backend. Duplicity uses the rsync algorithm to make backups, and stores the data that are backed up in .tar archives. Doing so, it does not matter what type of destination file system is used. All file information and unix permissions are stored within the tar archive. Thus, independent of where the tar archive is stored, all original data can be retrieved, including file attributes and unix permissions.
Incremental backups make for time and space efficiency
Duplicity makes incremental backups. That means that the very first time, all of your data are backed up and stored in a big initial tar archive. On subsequent backups, only changed files are included in the backup, and stored in smaller, additional tar files. These are the incremental backups.
That approach makes for a very space efficient backup. It explains why you see less data than you expected after making 5 backups. The same data is not duplicated in a next backup. That makes for space efficiency, but also for fast backups.
However, now and then, by default after three months, DejaDup will start over and create a new full backup. After a while, it will delete older full backups and the incremental backups that belong to it.
When you restore the data, the data in your original big tar file and the smaller tar files that captured the differences during subsequent backups will be used to reconstruct all of the original data. Optionally, you may restore just selected files, or older versions of the backup. These advanced features may not be exposed in the Deja Dup graphical interface.
Encryption
A little detail: duplicity also encrypts the data. So, apart from fully storing your data, it also keeps them safe from prying eyes.
