Agreed. I hope the guide goes as far as to start explaining how AES-NI can be utilized, and how constant time implementations can be written. I think the only improvements we can expect for AES (considering it's standardized) is wider deployment of proper implementations of it.