Forgive rambling reply from 71 year old toothless RURAL Appalachian (empathize with your power grid issues).
When qnaps with pcie slots became available i went to M.2 raid 1 system pools(hero) or system volumes(QTS). No noticeable difference on startup or shutdown times, but once up, QTS/hero gui is more responsive. Allegedly 5.2 was supposed to improve boot/shutdown times (Q college U-tube 'Introducing QTS/QuTS 5.2 security and performance improvements") but I MAY have noticed infinitesimal improvement only on latest models running hero.
Primary reason for this configuration is to allow the ability to safely detach and physically remove DATA ONLY HDD raid pools. This eases offsite storage of data raid pool live images. Before any firmware upgrades I ‘safely detach’ data raid pools and unplug/remove their HDD’s so I can shakedown upgrade with spare disks before trusting it with my live data (by re-inserting those HDD’s and doing ‘scan and recover’).
Regarding OS/system volumes, I BELIEVE this is how it works and defer to one who actually KNOWS. The raid 1 mirrors splashed across 5 or 6 small partitions on all drives are created when a new disk is inserted or read/updated when a used-Q disk is re-inserted. I believe these have relatively static info about hardware model, pcie’s, drives, os version and release level and I suspect pool and raid config and would only be accessed at boot and updated when any of those things change. You cannot see these from the QTS/hero gui. When doing initialize from scratch, when you first can get into the gui, you will find no pools/volumes/shares show in Storage and Snapshots. If you try to go into QuLog or install an app QTS/hero will tell you no system volume (QTS) or pool(hero) is available and needs to be created. Be aware that the very first volume/pool you create will be tagged as “(System)” in the storage/snapshots display, and the only way that can be changed/moved is with re-init from scratch. The “(System)” volume/pool has QTS/hero logs, applications and other more DYNAMIC QTS/hero OS data.
That and use of HDD spindown settings is why user noticed NAS going quiet moving system to ssd’s.
Even though never supported, I had your original proposed upgrade plan (with slight mod) work for me numerous times when first implementing M.2 raid 1’s as “(System)” volumes. Variation from your plan (assuming your 4 HDD Raid 5 data volume is ONLY volume in pool and shows as “(System)”) is to create a second volume for data and shrink original '(System)" volume as you move data to new volume and resize new volume absorbing freed space until all the data you want to keep is on the new volume. That process would proceed the steps in your procedure. After step 6 in your procedure, if you’re in luck when you attempt to ‘scan and recover’) you will attach an additional pool with two volumes, neither flagged “(System”) and now you can delete the one that WAS flagged (System) prior to your re-init on M.2 ssd’s.
If your in luck, that may save you from having to recreate your 4hdd raid 5 and restore data to it.
Be aware that while that worked for me a half dozen times or more when first going to M.2 raid 1 pool '(System)" volumes, each 4.x QTS upgrade seemed to get more picky about what it was willing to successfully ‘scan and recover’. In fact a procedure that is actually supported, failed me on one of seven 5.2 upgrades. On my TS-EC1280U-RP, like all the others I safely detached data pool and unplugged its disks prior to the firmware upgrade, but when I went to ‘scan and recover’ after the firmware upgrade and testing, it failed to recover the data pool. I knew QNAP support might be able to help but estimated the 10-12 days to recreate and restore 70TB pool might be faster than support so I went to recreate pool on same hdd’s and it refused to use them until I reinitialized the system and data pool hdd’s from scratch.
Hope that helps…