Science

Well its just silly. To me its demanding that no mechanism of free choice (will) be present.

Well.. yes, if that is the implication, then well.. that's the implication. But changing the definition of freedom just because one does not like the implications would not make much sense to me.

To me you can’t really include the individual and their molecules and such as they are the individual that has the agency.

The whole point is to find out if the individual has freedom or not.. and for that you need to find out if each and every of the molecules that make up that individual has been influenced by external factors past and present. Emphasis in past, given that there are way more external factors in our past than there are in our present. The external factors go even beyond our own lifetimes.. you are influenced by factors that happened before you were born.. and before your parents were born.. we have external factors even encoded in our DNA. The reason why you even feel the impulse to eat that donut is an external factor... the reason why you feel the impulse to NOT eat that donut is another external factor. Our individual wishes and wants are just the expression of external factors... our molecules are built through external factors, external factors make us grow.. external factors are what we are made out from. And since we all are modeled in different circumstances, we are different, every last one of us, with different combinations of different external factors, each influencing each other, in a harmonious soup of relationships with one another and with our environment. Even 2 identical twins become more and more different the more they experience the world, since they will inevitably have different experiences in life (even from the womb, before they are even born!)... even the smallest of things can snowball into experiences we will unconsciously internalize and alter our personality and the way we see ourselves and others.

Even though theoretically its considered the best way to go and even if the individual knows the optimized choice they may choose to not go that way.

I disagree. People always take the optimal choice for them. It's just that what's "optimal" depends on the dataset one uses. Eating that donut has a lot of pros, and not eating it has a lot of pros too.. it's all about what action is optimal given the influence of the external factors that model one's behavior. In some models, eating the donut and getting the gratification is the optimal path, so they do that (whether it's good for the individual or not), other neural models might see more value in not eating it so they don't do that.