MB$ vs UEFI would be a critical issue for what you’re doing.
Beware pointing VMs to raw disks, there are good reasons why that’s never recommended in any virtualization technology and only more experienced Users who understand the issues might do that.
The risks of disk corruption goes up considerably, and usually relate to more than one OS modifying files because the usual protections isolating running OS are removed. You must intimately know not only the obvious but also the less documented background processes running in every OS you run. The fact that your raw disks have OS installed on them and the disks aren’t simply storage increases your risks further.
In other words,
Don’t do what you are doing if you value your systems. Don’t do anything you can’t afford to lose.
Yep, I learnt that the hardish way. I say hardish because I have a data partition that all my distros read and write from extensively. The data partition wouldn’t mount on boot when I rebooted my real (as opposed to virtual) Tumbleweed system; fortunately it just needed a little fscking before it would, with minimal lasting damage. It was good in that there was no significant lasting damage, but at the same time it gave me enough of a scare not to do it again. Thanks, however, mate it’s helpful that anyone that sees this thread in the future knows the risk.