I spoke about the Windows problem existing on the 680 with Mac EFI.
Ok for the Windows bug I'll restore the Mac EFI to the card and make a video. Someone else will then confirm if the bugs exists on 7 and 9 series.
For the OSX bugs, I don't need to replicate that because it's already on 5-6 forums. It affects 6, 7 and 9 series.
I was following your post. If I understand correctly.
1) You have a EVGA 680 (PC version)
2) you flash it by yourself (that means the BIOS part is not originally from your card)
Therefore, you can't rule out that is the problem by mismatching BIOS. Even it was ok, but it doesn't mean that a BIOS from other card will always work after driver update, right?
What MVC suggest is that unless you can create a EFI ROM that 100% base on the original ROM, that "crash because Mac EFI" cannot be a valid argument. That's sounds very logical to me. But it sounds almost impossible to achieve for a normal Mac user like you and me.
No matter your 680 how similar to the official Mac Edition card, they may not be identical. So, I think there is another way to prove it. e.g. You open the EFI ROM, remove (or ignore) the EFI part. And compare the BIOS part with your original ROM (but by bit). If they are identical, then IMO, your "EFI cause crash" can be valid as well.
Or, if possible, remove the EFI part and then flash that possible problematic BIOS on to your card. To check if problem presist.
I never touch anything about Nvidia EFI ROM. Not sure if it's as simple as the AMD EFI ROM to deal with. If it is, remove the EFI part should be achievable.