At some point the size of the brain also becomes a risk during pregnancy, too, which is why infancy is a thing.
The heuristic 'X is the reason why Y' is a dangerous intellectual shortcut; Reality is typically FAR more complex than that.
When it comes to evolved biological traits, this mode of thought is wrong more often than it is right. The sheer number of degrees of freedom in biological evolution is vast, and effects rarely have fewer than dozens of causes; often have thousands of causes; and as far as I am aware never have a single definitive cause.
Science uses a hierarchy of levels of abstraction, so mathematics is provable and can produce definitive statements of fact; physics is an abstraction of mathematics; chemistry is an abstraction of physics; biology is an abstraction of chemistry; and evolutionary theory and psychology (of which discussion of 'intelligence' is a part) are abstractions of biology.
At each level, different tools are required, and the rules are statistical summaries of the underlying rules at the previous level. It's theoretically possible to model (almost) everything using quantum physics, but trying to do chemistry by calculating the quantum states of all of the subatomic particles would be practically impossible for all but the simplest interactions.
The claims you are making about intelligence appear to be founded in the simplifying assumption that the huge degree of complexity at the evolutionary level can be safely disregarded in favour of an heuristic that allows us to make simple logical connections between effects and their supposed causes. But we know that this is far less likely to be true at the biological level, and that while it's necessary to speak at the next level of abstraction beyond biology in order to say anything at all about the subject, its also impossible with the current state of human knowledge to do so with great confidence.
TL;DR - It's all a LOT more complicated than that.
Yes, I agree it must be a very complex thing.
Still, I think you are being a bit harsh with rousseau here. He didn't do what you seem clearly to suggest he did. He was talking of "a reason why", you are talking of causes. So, there's already a semantic gulf. Further, unlike causes, reasons are definitely things we know of and understand, and as such they are what humans can trade between themselves to come to a consensus as to what, then, might be causes. Sciences doesn't work differently, in that it's just the most effective process we know of to arrive at what we can agree are not only very good reasons, but our very best reasons, and reasons to arrive at a consensus as to causes.
I agree with rousseau that the size of the brain is definitely a risk during parturition and as such a limiting
constraint to the size of the brain in humans, and,
plausibly, a limiting constraint on human intelligence. I would assume that's a
reasoning accepted by the specialists.
And we're not even doing any kind of science here, so our reasons can only be a bit fuzzy.
And actual causes must be at quantum level, so we're very, very unlikely ever to tease them out. All we have are reasons to believe in causes.
EB