I couldn't agree more. Very nice description of the CLT. One other thing that is not often mentioned in this blog post or in many other places is that not only can the population mean be estimated from the sample distribution but the population variance can be estimate from the distribution of the sample variance. This one is not as straightforward, but is often quite useful.
