Whether you choose to dismiss the methods for booting I show, as I have reported in the past, in most every case, the methods mirror what is seen on bare-metal.
My above statement is NOT meant to anger, rather it intends to re-enforce what I have found in past years testing. The Ventoy -> SG2D USB allows me to first test in Ventoy, then to test in SG2D without a need to change USB sticks when validating the ability to boot. For all PUPs & OGs (excepting a very-very few) if the distro does not boot in Ventoy, I then check to validate behavior via SG2D in BOTH bare-metal and VM environmental tests.
Their more, of course, in the benefit of that USB.
I also have and companion test a SG2D only stick to verify when I observe misbehaviors in both Ventoy-SG2D USD and/or QEMU such that I can provide accurate feedback whenever boot-time issues surface.
In all cases, I try as best I can to give sufficient info to the developer that might be useful for any future movements they decide in an effort to be helpful.
7 methods to boot any and every bootable ISO file tested over the past 5 years. This includes IBM/Apple/Microsoft/Debian/ and other big box and small box offerings.
I do understand that boot issues can be frustrating to deal with, as many-many factors can affect the various things I've seen in distro boots over the years. Years ago, for me, it was the constant BIOS changes, then Windows came along and more BIOS changes, then Apple, then UEFI, then ... well we know this story as we, all, have seen them ... constantly.
I admire all developers as they have to endure the multiplicity of elements that they face constantly in the software world for the world of hardware their software must run on. This gets even crazier when you throw in the varied filesystem formats which can/do affect things as well.
Yes, it can be headachey, I agree.
Please dont let my post bother as it merely is just giving some examples of what I am seeing. The one benefit that QEMU afford me is that VM stanzas I use creates the EXACT Virtual PC no matter which OS I run on whichever PC is used: That virtualPC will always be an exact replica. sThis is beneficial in that as I investigate behavior in bare-metal I take note of the hardware differences between the Virtual PC versus the bare-metal.
After I complete a couple more promises I've made, I will, next, try as best I can to create a VM stanza that I am confident in which is a UEFI version. I feel the BIOS VM I currently report will be obsolete for comparison soon, if not already. To do so, of course, there are several issues that come into play as UEFI is one thing while SecureBoot is another as well as Legacy UEFI is yet another...you get the picture.
Please understand, that I am only the messenger in my reporting. It is merely my observations and does NOT mean anyone needs to change unless it is important to do so.