I don't have much respect for the school system's standards for what age is appropriate to start teaching different subjects. Children in my family start reading around 3. I consider reading a power tool that bootstraps further learning. If I had waited until school to start learning, I would have been held back in my development for years.
Children have wildly different capabilities. I've seen quite a few who probably wouldn't be able to read at age three. As the father of two highly intelligent kids myself (according to a few unprompted remarks by a number of people), I don't see what the rush is either. Kids are kids, let them play! They can be serious the rest of their lives. If they are intelligent enough to read at three they'll be fine anyway....
If your kids can read at 3, then good. But the power-tool argument allows for flexibility.
I think I was a little slow, or at best normal in picking up reading as a kid. But for whatever reason, by about the age of 8 I was reading well ahead of the rest. So I never read The Cat in the Hat, but I read the The Hobbit long before most kids.
And I'm glad, because that second level is where the real power-tool lies. There's little point reading non-fiction at Cat in the Hat level. But there's plenty of facts available that are simpler to read than The Hobbit.
That isn't to say I didn't learn other stuff: my favorite childhood book was about science. But it had lots of pictures, and when fist got it, my parents read an explained it to me.
I was a precocious reader (<3yo), and I have a strong suspicion that kids who learn to read early read differently than those who learn "on schedule". I think early readers do something much more like code breaking, and I wonder if this affects second language acquisition or other skills. I don't necessarily think one style is preferable to the other, but I do think there is a potentially interesting difference.
I was very happy to read early, and I think it was advantageous for me, but I have not seen evidence that early reading in itself provides any lasting advantages on average. There may be a good deal of heterogeneity in the early-reading population.
"Code breaking." Interesting. I think I agree. I learned to read around 4 years old, and I did so by trying to follow along and predict words as my dad read to me. I can recall thinking "ok, there is a word coming up that looks like 'be'" and then verifying it was correct when he read it. I taught myself in that fashion. I can recall passing a road sign and reading "tippecanoe" aloud and surprising my parents shortly thereafter.
When we taught our own kids to read, I was surprised that this was not the way that they would learn to read. We ended up having good success with hooked on phonics with our first two and the last one did well with vast amounts of repetition and memorization of his favorite books.
