IacopoV:
I think you may be confusing equivalence with identity. In particular, when you discuss two brains in the same state, you're talking about two brains that are equivalent. But if you're talking about two brains at all, then you're automatically not talking about identical brains. In a lot of your setups, you're establishing two brains as equivalent, and then treating them as if they are therefore identical; as if, given X is equivalent to Y, everything true of X is true of Y.
That's not exactly how it works.
Let me illustrate. Suppose I clone you this time. To alleviate confusion, I'll clone you twice, making both clones simultaneously, and then will destroy the original. This way, we remove all arbitrary reference points such that we can't possibly label one more authentic than the other. So let's suppose these two clones start out as equivalent as you would like to make them, and I put them each into respective laboratories that are antipodal to each other, located on the equator. Let's call one of these clones W. The other, we'll call E.
Now we'll carry out an experiment. I'll grant W a subjective experience; let's say, I give W a chocolate brownie. Meanwhile, in the other lab, I'll use exactly the same environmental conditions, giving E exactly the same chocolate brownie experience. Maybe. Or, perhaps I won't. It'll just have to depend on my mood that day.
So let's focus on W. While W is enjoying my chocolate brownie, I'll pose a few questions to W.
1. Do you know you are eating a brownie?
2. What does it taste like? Is it good?
3. Do you know if your clone is also enjoying the brownie? (Or hating it, as the case may be?)
Now, here is what I would expect. I want to know if you would expect the same thing. To question 1, W knows that W is eating a brownie. To question 2, W probably enjoys the brownie (I really do make a killer brownie); though I leave open the possibility that brownies just aren't W's thing.
But to question 3, W has no clue if E is enjoying the brownie. So now we have to talk about two possibilities.
Possibility A: E is being subjected to the same exact questions the same way, and is equally unsure that W is enjoying a brownie.
Possibility B: E is experiencing different brain states. No brownie. Possibly E's enjoying a nice scone instead.
Here's the problem. On the off chance that possibility A is correct, W's brain state is equivalent to E's brain state. Nevertheless, W cannot tell the difference between possibility A or possibility B, because W is not privy to E's experiences. However, W does know what W is experiencing.
So even though W and E's brain states may be equivalent, W and E are not identical. E knows something W doesn't know (namely, E knows what E is eating), and all it takes is for there to be a break in the causal chain between E's brain states and W's.