It is only for new data.

For example, you would have to defragment your filesystem again with btrfs filesystem defragment -r -v -czstd /. Where zstd is an algorithm and /, a root path. With this command, the default compression level will be used, which is level 3.

Be careful, defragmenting the btrfs file system will/can duplicate the data.

As for a mount point, if you decided to use zstd algorithm with level 1 compression, just add the compress=zstd:1 or compress-force=zstd:1 to the mount options (fstab or while mounting manually)

No, but according to this Phoronix article, they will fix the RAID56 issues soon:

The support for RAID56 is in development and will eventually fix the problems with the current implementation. This is a backward incompatible feature and has to be enabled at mkfs time.

It generates initramfs images faster. Also the generated images are much smaller (30-40MB for fallback images).

The only problem is that the tpm2 unlocking doesn't work currently.

If I know correctly, defrag will always duplicate the reflink files.

https://btrfs.readthedocs.io/en/latest/Defragmentation.html

Defragmentation does not preserve extent sharing, e.g. files created by cp --reflink or existing on multiple snapshots. Due to that the data space consumption may increase.

It is fine. You can use the duperemove tool (or bees) to find and remove duplicates.

https://btrfs.readthedocs.io/en/latest/Deduplication.html

So it is out-of-band deduplication and has to be done manually.

Also, by default cp and most file managers use a reflink copy (data blocks are copied only when modified)