Thanks for the explanation, cantata!

First off, it´s been a long time since I have read Kant, so I really don´t remember much. I am aware that this is but your interpretation of it (albeit one that seems to show that you have spent quite some time and effort in understanding Kant - so I have no reason to doubt that it is a good summary).
Anyway - we´ll both keep in mind that I am only addressing your interpretation, and of course only in the way I manage to understand and interprete your interpretation, ok?
I am trying hard to keep my skepticism towards the idea of "freedom" out of it - I am trying to concede for the sake of the argument that there can be such. I don´t know exactly what Kant means when saying it, but for a formal investigation of his argument this isn´t even necessary.
But even when conceding "freedom" were possible, and even if conceding Kant has some sort of clear concept "freedom", I still don´t see the logic in his argument:
Yes, I think it is circular. We would need to be (and/or feel free) in order to decide to act morally, in the first place. So we need to be free to act morally (in the way Kant defines "moral acting" and "freedom") and we need to act morally in order to be free.
Next: I don´t see how acting morally is necessary to gain this (feeling of) freedom. The mere idea that I can act morally or immorally would be enough. Even if I acted immorally on purpose and in full awareness of the immorality of my act, I would still exert this freedom.
The dichotomy "acting upon desire" and "acting morally" is a false one, imo. I could act against my desires and still purposefully immorally, after all.
But even if I acted upon my desires (while disregarding my moral ideas) I would still exert my freedom to act upon morality or not.
Another big issue for me: I don´t understand why Kant considers acting upon reason (or morals, which seems to make no difference to him here) an act of freedom, but not acting upon desires - other than that he would like it to be so. What I suspect him doing here is: Hiding the desired conclusion already in his premise, defining words so that they help his case, and so making a great intellectual fuzz without actually getting anywhere. Reason/freedom/morals - they all seem to be by and larged used for the same concept, and he uses them to define them mutually, and as a result they all appear to be prerequisites for each other.
Now, I understand that he values reason, freedom and morals high - but - if I am not missing anything - he could simply say so. His complex deductions don´t seem to add anything of significance to that.
But may be I am indeed missing something.
Maybe playing the devil´s advocate will help sort things out:
I postulate that in order to be free we have to act upon our desires. They are what make us humans, they evolve within us, they are expression of our personalities. We have to escape the prison that morals and reason are for us. Acting upon reason would determine each of us to act the same way, after all, so it can´t lead to freedom.
I have tried to use the same argumentative method that Kant used.
You have to remember that Kant doesn't try to prove that we're free. He only says that we need to suppose that we are if we're to get anything done (a proposition that I know you disagree with - and one which rather causes a problem for our hypothesis that language is part of the reason that most English-speakers have a concept of free will, and you don't, since of course Kant was German

). For Kant, the
incentive to act morally is that it is only when we do so that we can be free. He takes it as a given that people want to be free. Now if you don't believe that such a thing is possible, or indeed, if the notion of being free doesn't appeal to you anyway (and I should imagine that, like me, you find the idea of it too confusing to be attracted by it), Kant's argument in favour of acting morally holds very little water.
Now, that said, I believe he has what he thinks is a sound reason for regarding reason as free and desire as determined, although it's probably going to sound circular when I try to explain it. Please note that this is just my interpretation of Kant; I may have him all wrong!
Kant has a rather Cartesian view that what makes us human is our ability to reason. So he then imagines what purely rational beings would do - that is, beings which inhabit the noumenal world, the world of things as they actually are. In the noumenal world, we just don't know how things go on - and we certainly don't know anything about the role of causation. You can transpose in here Hume's scepticism about causation if it will help - like Hume, I believe that Kant thinks causation is an appearance in the phenomenal world, but not necessarily an actuality in the noumenal world. He comes to the conclusion that perfectly rational creatures would act according to the categorical imperative ("Act only on the maxim such that you can at the same time will that it become a universal law") - in other words, they would always act morally. And rational creatures who inhabit the noumenal world are the only creatures with a
hope of acting freely, since they inhabit a world in which we
don't know whether causation behaves as it does in the phenomenal world. Therefore in order to escape determinism, our only chance is to apply our reason, because our reason might not be bound by determinism.
There's no need to imagine anything weird about the noumenal and phenomenal worlds - the former is simply the world as it is without the hindrance of the veil of perception. I know that you and I have different views about the nature of reality, and I suspect this might also be a sticking point between you and Kant.