I've become aware of the issues with APFS formatted spinning disks - a performance hit - confirmed by others probably also related to RAID formatted APFS spinning disks. Maybe you could confirm if an issue with RAID formatting of APFS spinning arrays in your testing and how big a hit there is between APFS and HFS+ on spinning disk arrays?
I tested two identical drive models, a single disk on a Mac Mini 2023 formatted one as HFS+ and the other as APFS Encrypted. I read about a difference even with APFS (without encryption) due to the ID track location and spinning disk in APFS.
Macbook Pro 16" Retina XDR, 2024 M4 Pro internal 2TB storage, 36GB RAM, MacOS 15.6.1 Sequoia, running v8.0 SoftRAID software; Local RAID drives/enclosures: 4M2 OWC Enclosure with 6TB NMe, RAID4 Storage; two external OWC T3 enclosures (2.5TB online storage) populated with JBOD 6x500GB, EVO SSD, RAID 4 array disks/partitions; Local Backup: 2TB.
Once you understand how APFS works, you will see why it is not optimal for HDDs.
Every time you modify a file, APFS saves the edits/changes in sa separate segment. So a file quickly becomes comprised of thousands of segments, even with modest edits.
This is not really a problem with SSD's. With HDD's, its a massive performance hit, that gets worse the more you use a volume.
If Apple were to implement a feature in APFS, called copy on write (meaning it would not fragment files, but write the entire file out, similar to how HFS saves a file), then this issue would not be a problem.
It appears unlikely this feature will ever be implemented.

