(Extra bits edited in follow:)
I detect another implied point, which is that there are levels of ability to reason. That is, by analogy with cells, which are dumb, inside a human, which is smart, the supposition is that humans are relatively dumb components inside a universe which is super-smart. This relies on the concept of super-smart having a meaning. (It also implies that the cells are less than totally dumb.) Then humans are somewhere on a sliding scale between totally dumb and infinitely smart. But I see no reason to make this supposition that such a scale exists or has any meaning. So far as I can see, there's only one kind of reasoning and it doesn't have levels.
I guess there's things like squirrels solving puzzles to get nuts out of a container. But I think that's a different function from understanding stuff.
We have managed to hack our own limited intelligence by using collective memory so we aren't starting from zero every generation, but we still aren't naturally inclined to come up to things we readily understand at our scale and grasp what they might represent on much smaller or larger scales than our own. Even for people trained in these fields it is extremely challenging due to the problems with scale and our frame of reference. We had to develop things like microscopes and telescopes to take objects small and large and either magnify them or reduce them to something we can actually begin to make guesses about at our scale.
This seems separate from capacity to understand them, after gaining the ability to observe them.
I'm not convinced by the second part, about human knowledge becoming too gnarly for humans to cope with except by group effort. This a breadth vs. depth question, but depth is the winner over time I think.
