One problem with this is how does true randomness creep into a deterministic system of choices made by living beings? Studies in entomology have shown that insects, even when the environment is very limited and variables strictly accounted for, will still move in individualistic, unpredictable patterns.
It would be quite possible for parts of the neural network to be sensitive to random influences, such as noise in the system, or even generate random (or pseudo-random?) noise, and I would be surprised if simple organisms didn't make use of randomness in their tropisms. It's also possible that this kind of thing is used in higher animals (some sensory processing has been shown to use noise to raise the threshold of sensitivity). But the main point is that randomness is not necessary if the output of the processing is sensitively dependent on the initial (and ongoing) states of the processor and/or if self-referential processing is involved. Remember that this is not a passive system simply awaiting the inputs on which to make a decision, but a complex active system, with numerous continually changing states.
As for free will in more complicated beings like humans, I think you have to take things like cognitive dissonance into account. You make it sound as though a person's mind will constantly weigh out biases and information and come to a singular choice of absolute resolve, but I think you are oversimplifying it.
You inevitably have to simplify these things when describing them in a paragraph or two.
For example, the conscious mind often has a very different view of a given situation than the unconscious mind. Sometimes I make decisions I absolutely hate, knowing full well that I should not choose a given path even as I am making the decision, but yet I allow myself to cave to my unconscious fears. These unconscious fears are not just simply biases or prejudices; they are things my conscious mind absolutely 100 percent knows are no big deal and yet I sometimes allow them to have power over me. How can all decisions be described as a deterministic weighing-out process if there exists more than one part of the mind capable of making decisions?
You are quite right that there seem to be many sub-processes or sub-modules involved, each with competing priorities, and they all provide input into the evaluation. At some point an action must be initiated and it seems to me that it will depend on the status of the system at that time - crudely, which of these competing priorities has had the largest influence during the evaluation. It seems likely that preparation for one or more potential actions builds up as the evaluation is in progress, so that the appropriate behaviours can be initiated for the action that is best prepared for - just my speculation.
If you say that one side of the mind will always triumph over the other, what causes the success in light of the fact that on another day with a similar mindset, the other path may have been equally likely to occur? Sure, it could be a complex illusion to believe that both paths are equally likely to occur especially given so many variables in human life, but I don't see why if it is an illusion, the illusion should exist in the first place.
I don't really see the problem. It doesn't really matter which of the courses of action is selected or which of the internal modules 'wins out', the result is what we call our 'choice', and if we feel it was uncoerced, we call the process that gave that result 'exercising our free will'.
The system has evolved to resolve the competing priorities in a timely manner so that the most appropriate behaviours can be initiated. It's not necessarily a rational process, and the contribution from introspective conscious awareness complicates matters and can make it dangerously unreliable (although the benefits must outweigh the disadvantages, or we wouldn't be here). The system is arranged so that, in general, sufficient processing and evaluation is done to select appropriate behaviours, and these behaviours are initiated while they are still appropriate - we wouldn't survive long if we didn't think enough about what to do next, or if we thought so long that we missed our opportunities.
It seems to me that the nuts & bolts of this processing is below our conscious awareness - we will be aware of some of the competing priorities, some idea of the leading preference, and we can guide the focus of attention to particular areas, but there is evidence that the evaluation is completed and behaviours are initiated subconsciously, and the 'self' becomes aware of this after the fact and a plausible narrative is generated to explain it. It's still 'us' exercising 'free will', but we're not necessarily consciously aware of all the details of the decision making process, any more than we're aware of how we generate our utterances in conversation. Who hasn't blurted out some tactless comment or made some particularly clever quip and wondered where it came from?
If free will wasn't real, why would the mind need to create unpleasantness as a side effect for a decision we know we probably shouldn't have made? In this case, we should just be philosophical zombies who would experience emotions as mere attachment with no sense of "qualia." From an evolutionary standpoint, we don't need to have such feelings of dissonance to function and be successful so they should simply not exist because they actually get in the way of our pursuits.
I don't particularly see what unpleasantness has to do with free will. From an evolutionary viewpoint we do need to distinguish between what is desirable and what is undesirable, and the senses and the sensations they produce are the means by which this can be done. Unpleasantness crudely steers us away from things that may be detrimental - it's a fundamental principle in the evolution of living things. Without pain we wouldn't be aware of damage and learn to avoid the things that cause it. Without hunger we wouldn't eat, without fear we'd get into too much trouble. The mechanisms that manage these sensations are inherited from our earliest beginnings and haven't changed much in millions of years. Our fancy brains are built on top of these foundations.
As another example, if I had a nightmare last night or even daydream that I found extremely unpleasant, what allowed my unconscious mind to allow for something that my conscious mind would have vetoed?
Why do we dream? Why do we have nightmares? I don't think anyone really knows for sure, but in sleep the confabulator or narrative generator is allowed to free-associate with minimal constraints of logic and commonsense, perhaps to make connections between recent experiences and memories, or to produce new links between apparently unrelated areas (e.g. creativity?). Given the importance of unpleasantness to survival, it's not surprising it can appear in dreams, if dreams have some constructive purpose and survival value. Personally, I only seem to get nightmares when I'm too hot in bed - as if the subconscious uses the physiological effects to wake me up so I can throw off some bedclothes...