It's been years since I've taken a neuroscience course, so please forgive by lack of technical jargon.
There is something that I remember from these courses that my prof. referred to as 'windows of opportunity'.
Essentially, the brain develops at a certain rate of progression, and some skills with varying levels of abstraction, are better learned at certain times.
For example, languages are best learnt young. Most (not all) people who learn a second or third language after the 'window of opportunity' closes, will very likely always have an accent or have something else that differentiates them from a native speaker.
The same goes with teaching someone something when they are too young. The example that most clearly sticks out in my memory was the teaching of grade-school children more complex and abstract math (Junior High level). The children were able to learn the math eventually, but years down the line, they plataeud at a much lower level than their peers who learned at the regular progression.
Granted, that said research is not currenlty out-dated, what does the neuroscience community think of this popular "my baby can read" product? While it may work as promised, is there any chance that learning to read too early can ultimately be detrimental? Any long term studies on this product out yet?