This seems like more work than it is worth. Why not use mdadm raid for the boot volume? You can use the standard Debian installer to do that and then install OMV on top of that. Very simple. Plus, if the motherboard fails, you could move your setup to another computer. If you use the b120i for raid, you would have to have another b120i based motherboard to recover.
Personally, I don't think the boot volume needs raid as long as you back it up once in a while (backup plugin and cron job).
OK, then I'll try using the B120i in AHCI mode, separating the two SSDs. And then maybe use the secondary SSD to rsync to every night or something. But then I could perhaps also have OMV do software raid-1 on those two , or am I missing something? Maybe that's just a waste of SSD writes, which I also consider to be a thing now. (SSDs do fail, I've already seen a couple of Intel SLC ones that were allegedly deemed indestructible for 50 years, not so alas..) And of course electricity. I could just have the extra SSD spin down until it's used for the delta-backups every night..
Motherboard failures are so rare to me with HP machines, I must say I can't even remember it seeing once under my hands. I did have some Apple, Intel, Gigabyte, Asus, MSI and AsRock boards fail on me in my 30+ years computer career, but never Compaq or HP. But that's just my experience. It's the reason I went for this Gen8. Where I work we have many HP servers deployed in office as well as in data-centers, they're solid as a rock.