This is similar to Marx's critique of freedom under liberalism as merely 'formal'. The problem is the gap between that can exist between a nominal right and practical exercise of that right.
This kind of problem is common with rights-based approaches to justice and can be witnessed with human rights broadly. Its identification isn't unique to Marxism, either; liberals sometimes get at it with the phrase 'equality of opportunity', for example. To say that opportunities can be unequal (and that this is a problem) is to admit that justice requires the guarantee of more than just formal rights. I'd say this a problem that has shaped liberal 'privilege' discourse as well: privilege is just such a kind of gap that allows (or constitutes?) the persistence of injustice in the face of nominal/formal/legal equality.
Like in other cases, I'd say that the four fundamental software freedoms get at something genuinely important, and that it's better to have them, even just formally, than not. But like with other freedoms and rights, it's easy to conceive of them too 'thinly'. They need to be fleshed out with a more general awareness of power relations and of the practicality of their own exercise.
To some extent, that awareness of software freedom as situated within power relations is actually already present in free software discourses, which talk often of things like subordination, domination, subjugation, etc., from the start. Unsurprisingly, that dimension is largely absent from the 'open-source' perspective.
It might do something in humans, but the idea that autism is reducible to genes— and a single gene, at that— strikes me as laughable on its face.