Cool discussion going on here. What I just wanted to add is that the actual mechanism of M1 and M2 and their transition is not fully (!) researched. There is quite some research on those modes, but basically all of those studies suffer from one of the following problems:

- Tests mainly done on untrained subjects (not managing the passaggio)
- Tests mainly done on classical singers (fixed sound ideal)
- Very sparse material on smooth (!) transitioning between the two

What is quite clear is that M2 has a more shallow vertical phase in vibratory pattern. Only the outer layers of the folds are in vibration and have contact. This is the main difference between the modes. Additionally, without "techniques" M2 tends to have a lower closing quotient compared to M1, but this can be changed dramatically by technical adjustments.

However: some experiments (like the latest CVT ones) indicate that there can be situations of a "reduced vertical phase", which would actually be something like an "in-between" state between the two vocal modes.

And just another word on the term "mixed voice": An interpretation of that term that is still valid is if you consider it "mixing of natural properties". Just think about it like this: The vocal mode for the high notes WILL be M2 at some point and for the low notes it WILL be M1, so you could say that the use of M1 is "natural" in the low range and M2 is "natural" in the high range. Additionally M1 tends to have a high adduction and M2 tends to have a low adduction. M1 tends to have a big vertical phase, M2 has a low vertical phase. So the term "mixing" could refer to one of the following actions:

- bringing adduction (typical property of M1) into M2
- bringing a lowered vertical phase (typical property of M2) into M1
- lowering adduction (typical property of M2) into M1

... and so on, these are all different possibilities to "mix" mode qualities.