Disk Not Ejected Properly

Not a direct answer to the question, but your disks are configured as a striped raid (known as RAID 0).

This is not the best setup for a backup drive. It's often joked that the '0' in RAID 0 stands for 'no redundancy', because that's what you have.

Data written to this volume is striped across the drives - half to one drive, half to the other. The advantage is speed (each drive only has to do half the work, so things are (approx.) twice as fast).

The downside is that no drive has a complete picture of the data. This means losing either drive will result in the loss of all data. Hardly the ideal setup for a backup solution.

RAID 0 is best suited for performance, where the workload is spread across drives to improve throughput, or for capacity where you need the most space possible and don't care so much about data loss.

For backup purposes you should consider RAID 1, which mirrors both drives, so data written to one is also written to the other, giving you a better chance of recovering your data when (not if) one drive fails. Of course, your capacity is limited to the smallest drive in the RAID group (in this case, 28TB).

With Time Machine, though, you'd actually be better off running them as two separate drives, not as a RAID set. When Time Machine has multiple drives configured as the backup destination, it alternates the backups (backup 1 goes to drive 1, backup 2 to drive 2, backup 3 to drive 1, etc.). This is potentially superior to RAID because each drive has multiple distinct snapshots and if one drive fails, TM continues to run on the other drive, only losing the specific snapshots on the failed drive

For the specific issue - has this always been the case? or did it start recently?

My first thought would be power, but they're powered separately from the host system, so that's not likely to be a problem.

The second would be that one of the drives is taking longer to spin up on restart/wake, so the RAID set isn't 'ready' when the OS is trying to mount the volume. This could be an early warning sign that the drive is having issues, and your data is at risk.(see my note about RAID 0, above)

I occasionally experience the 'Disk Not Ejected Properly' message with my NAS that I use as my primary TM backup. I just dismiss the error, the drive reconnects automatically when the next hourly backup starts.

You could consider a 3rd party app that automatically ejects your mounted disks at sleep, such as Jettison.

I would echo the comment from Camelot that RAID0 is not optimal for a backup disk. A commonly recommended strategy is 3-2-1 where you have three copies of your data, two of which are backups and one of which is stored offsite. With your drive as RAID0, you have 2-0 (one in-use copy and one backup copy). My NAS is configured as RAID1, and my secondary TM backups are a pair of 4 TB SSDs that I connect weekly and swap offsite (so technically I'm 5-4-1 but overkill is better than the converse, IMO).

Many users have reported this issue when using sleep. The best cure is to not use sleep. Just have display turn off after a set time.