Ok, good. So, you’re blackout drunk and have no free will at the time but your brain is still making decisions and actions on your behalf. So if the conscious self is just awareness of the decisions and actions that the subconscious is making anyway, consciousness doesn’t equate to free will.
Part of your brain is yes - that happens when you're conscious too - you don't consciously decide when to breath or when your heart should beat right? I mean you can choose to take over your breathing - choose to hyperventilate or hold your breath just because... by your own free will, and if you want you could do it to the point where you pass out - though then you'll likely just start breathing automatically - which is just part of your brain running on autopilot.
If the predictions are 100% accurate based on deterministic analysis, then it suggests I have no agency to do anything different, and hence no free will.
Well, it wouldn't be 100% but for the sake of argument no it doesn't suggest that at all. I mean we can predict some behaviour broadly anyway - predicting some chess move doesn't remove agency... Some theoretical observer outside our universe might be able to predict everything even I mean I commented on that already so I'm not sure why introducing some constraints and talking about some theoretical finite, localized prediction adds anything beyond some theoretical observer that can predict literally everything?
It comes back to the point that we simply don’t know enough about our brain activity yet to say one way or another. It doesn’t have to be ‘magic’ or a ‘soul’ but if there is some element of unpredictability then it allows for the potential for free will, even if we don’t understand the mechanisms behind it yet.
I don't see what that adds here - like you say we don't know enough about our brain activity yet so why have a hard view re: some randomness or unpredictability being significant here?