I have often claimed on this blog that with regard to literacy, early is not necessarily better. Over and over again, we have seen autonomously educated children who didn't start to learn to read until they were 8 - 10 years old, rapidly progress to adult level reading and become avid readers.
I am now pretty sure that the same process applies to mathematical concepts, if usually slightly after the literacy competences kick in.
Apparently HEors are not the only ones to have observed this phenomenon. Now if only we could convince people like Zenna Atkins.