I agree with the others (and re 22B's point, I once posted here about a difficult conversation I had with a friend in which I contradicted her assertion that I couldn't tell he would still be ahead of his peers in x years (at school change) by pointing out that to be still ahead he didn't have to keep up the same, or any positive, pace - he only had to avoid going a large step backwards).
But... plateau can mean a lot of things. Beware the self-selecting sample here. I think it does happen, sometimes, that young children seem very ahead (perhaps because of an exceptionally enriched home environment, perhaps for some biological reason, who knows), and then blend back into the average. But their parents don't stay here, so you won't hear from them!
DS's learning patterns are certainly uneven sometimes. It's easy to believe that someone who saw only part of what he did could think he'd plateaued when his attention was elsewhere.
I fear that it's also possible for children with huge potential to never realise that potential, if they don't get good enough opportunities. By the nature of that, I can't prove it by anecdote. But I do wonder how much of "they all even out by 3rd grade" is caused by such a phenomenon.
