[Question] How to specify clone sources from snapshots of other unrelated subvolumes?

In the configuration `incremental_prefs`, all options either requires matching snapshot_name, or a parent_uuid relationship. How can I let btrbk use snapshots of other unrelated subvolumes (different name/path) as clone sources? This is useful when they actually share many data because of `cp --reflink=always` or [dedeplication](https://github.com/Zygo/bees).

The docs of `incremental_prefs`:
```
sro,srn
All from snapshot_dir matching snapshot_name, with parent_uuid relationship, sorted by btrbk timestamp (o=older n=newer).
sao,san
All from snapshot_dir matching snapshot_name, sorted by btrbk timestamp (o=older n=newer).
aro,arn
All from incremental_resolve, with parent_uuid relationship, sorted by cgen (o=older n=newer).
```

My use case is like this: I have two subvolumes `active` and `archive`, with different backup strategy and snapshot period. Once a while, I move inactive/immutable data from `active` to `archive` via `cp --reflink=always && rm`. To avoid needlessly transfer and/or store two copies of data on the backup device, I'd like to specify clone sources of other subvolumes when sending one, like `btrfs send -p archive.old -c active.old archive.new`. How can I do this in btrbk?

I tried `incremental_prefs defaults,sao,san,aro,arn` as suggested by documentations, but it does not seem to add cross-subvolume clone sources.

BTW: I'm not sure how to check what command line is used on send. `btrbk dryrun` does not print the btrfs command line to be invoked. I checked these by ro-mount the target device and view the error output from `btrbk run`.

Start with the incremental_prefs documentation and the btrbk dryrun/send path, then reproduce the active/archive case using unrelated subvolumes and reflinked data. Determine how clone sources are currently selected and whether the invoked btrfs send command exposes them; done means cross-subvolume clone sources work for this case and the command can be inspected.

Written by the indexing model from the issue text.