Correct way to initialize "parallel" restic repositories

I'm in the planning/experimenting stage of a 2-3 USB hard disks plus S3 cloud storage backup/archive solution.

Planned usage is as follow:

• The first restic repository will be initialized with past rsync archives/backups.
• That first repo will then be cloned to the other disks and to the cloud.

Once all clones are available:

• I will periodically add new backups on any available repo. Not always the same. Deduplication must properly kick-in.
• I will periodically copy snapshots from one repo to another. From any to any. Again, deduplication must work.
• And I obvously don't want to put the integrity or the security of the system in jeopardy.

Question is how to perform the clone of the initial repo.

• The most basic way is to rsync the folder from one disk to the other. This is bad, because the repo ID stays the same, which means that the local cache can't be trusted.
• Next option is to just initialize them independently, and populate the clones with restic copy. While this work, and following restic copy deduplicate properly, my tests show that a direct restic backup will no deduplicate.
• Next option is restic --from-repo original --copy-chunker-params. However, the design documentation hints more than once that the randomness of the chunker is part of the solution's security. I'm therefore assuming that reusing those parameters reduce the security.

Is there a better option? What's the exact security impact of copying the chunking configuration?