Which is
maybe normal, since there is a note about:
ROSE-storage - package adds additional enterprise data center functionality to RouterOS.
Current manual page does not include additional features included in ROSE-storage package.
The issue remains, however

, since there is seemingly nothing about spare disks in the Rose storage help page:
https://help.mikrotik.com/docs/spaces/R ... SE-storage
Nothing about the possibility of adding a spare disk, nor anything about a specific procedure to replace a failed drive

, I presume that the failed drive (or one "forced" to fail via raid-member-failed) needs to be removed and a new one added (hotswap or not?) and then either the array is rebuilt automatically or a raid-scrub needs to be initiated manually.
It is entirely possible that the "raid-role", defined as:
raid-role Defines the role of each device within the RAID array.
could be *somehow* used to identify a spare drive, but that in the examples posted is just the numbering of each drive corresponding to the physical slot (to make it easier, cleverly numbered starting from 0, whilst nvme slots are numbered starting from 1

).
Normis was excited by all this and mentioned "any number of hot spares":
viewtopic.php?t=214539#p1134077
any number of hot spares to automatically rebuild parity/mirror disks upon failure
but seemingly the good MIkrotik guys (mis-) writing the documentation lost the note about it.

(probably they placed it in the same drawer as the commands to execute on Client in the example to Mount a RAID array on another Mikrotik device

).