ok thank you.

Yes you can. you just need to manually (from the desktop) unmount the volume first, before replacing the disk, as the SoftRAID application does not yet know how to work with APFS volumes. But the driver will do the right behavior.

I have just created a raid 5 in 6.0.5 non beta version. Does this mean if a disk fails I wont be able to rebuild the damage disk?

APFS appears stable, yes. We are working on full application support for APFS, which is why we only support creating APFS in beta versions.

However, keep in mind two things:
APFS has no repair tools if there is damage to a volume directory

APFS is much slower file system and especially with HDD's can quickly get 50% slower than an HFS volume.

In 6.1, we should have full APFS support, the application change is not in the beta yet, but coming soon.

Notes on your comments:

You cannot certify an "APFS" disk, a certify is a raw mechanism test. There is no file system during a certify. Maybe what you mean is the inability of the application to unmount an APFS volume, so that the disk is released for a certify? If you unmount the volume first, then you can certify a disk.

APFS on windows was a big deal. We have that now, but only for RAID 1/0. More is coming.

I do not understand your write speed issues, I feel it has to do with Samsung. the SoftRAID driver will push any drive about as fast as possible, withing a few percent of its abilities. (RAID 4/5 parity calculations on slower CPU's excepted) I have seen Sameung blades perform slowly in the office, until they were certified again. I have read threads on places like tom's hardware and anandtech complaining about Samsung write behavior. But I have not researched this.

Thanks for the support and good luck!

The only way to certify a APFS disk in any of your apps, is to attempt it on APFS "Friendly" software and a hardware environment. That is in Big Sur and using SoftRAID 6.x. Even there, it proved dicey last time, so I see no reason to re-attempt it until there is a compatible software version that adapts SoftRAID across all existing platforms. What I have now is a work-around that barely does what I need with beta software. One you admit to be (and I have demonstrated as) full of holes.

When you have something that works in a current Beta version of Big Sur and on a release version of Catalina, it probably will need a complete reboot and reconfiguration anyway. So, let me know if that's available to test, I'll be happy to look into it further. 

My next step is more likely to be to put everything back to 5.x and HFS+ once again giving up on a failed experiment of SoftRAID6. That was the last stable configuration (which, BTW, as previously reported) one that had the same speed results as reported.

I would have to discount the equivalence of Write and Read speeds you claimed, as that has NEVER been my case. Not that SoftRAID isn't worthwhile for speed improvement and large storage. It's just not nearly as you have claimed with my Samsung SSD's and use case - on MBP 2019 and T3 (OWC brand Thunderbay 4) connected enclosures. 

You've got a ways to go, brothers (or sisters). Best wishes and happy holiday. 

Will look into it as time permits. I don't want to start over again recreating the volumes and retransferring the  data until your software is performing properly for my hardware (Intel). Regardless of the new struggles with recent hardware, there needs to be better operation with APFS than there is now on Intel devices. I think you've already been informed of the issues there, so far. As I see some movement in that regard, I'll make further decisions as to how to proceed. 

I think you need to certify the disks and create new volumes. This may simply be an SSD issue.