Is it reliable though? For example will the result always be exactly the same as if the drive was managed in Dom0? Does the used file-system matter if applied through a VM or are any possible factors completely unaffected? For example is there a difference to manage BTRFS, ZFS, NTFS, EXT4, random exotic FS, RAID of any build kind, HDD/SSD, old or new cutting edge drive technologies, or any other possible factor, through a VM compared to a bare metal OS?
Anything at all to look out for if undertaking changes on a drive through a VM?
Does it pose a risk change in the rate of bit errors? For example from file system or drive error odds with an 1 in 10^15, to 1 in 10^7 risk increase?
It is after all better to be proactive to prevent possible issues, than it is to leave it unknown, when it comes to precious irreplaceable important data.
Best practice in terms of security, yes sure, but is it also best practice in terms of data integrity? Is there any possible trade-offs here to be aware of?
There may be no difference at all, or there may be. But the point is, for those not in the knowing, which one is it? It would be great to be reasonably certain when using new technology where important data is involved, where common sense may need an update, rather than being among the first victims due to outdated assumptions, relics of an old age in an ever faster changing world.
So taking all that into account, all possible factors included, is it really just as reliable to manage drives in VM's as when done in Dom0?