There are a few sample questions for the NNAT in this file ...
http://harcourtassessment.com/NR/rdonlyres/F79BD2C3-574D-4C3E-AFF3-00A3BBD48113/0/NNAT_TR.pdf

I do get the feeling that this identifies kids that tend to be gifted more on the math/logic end of the spectrum. If my son went through a full IQ test, I'm sure he'd score less strong verbally, than logically and mathematically. He is very strong spatially. He's been doing lego sets for teenagers since he was 4. My husband and I both have multiple technical degress, so it seems logical.

Talking to a few other parents, I know that there are reading kindergarteners who were not identified at the 92% level on the NNAT. However, most of them also had fairly rigorous all day preschool programs, so maybe that is a factor as well? Especially some of the Montissouris around here seem to really push reading for Pre-k when the kids do full day.

And I do agree - it is totally possible for a kindergartener not to be able to do their best on this! It seems like a useful tool for diverse school districts as a baseline. Our district does testing at the 2nd grade level also, and at any other grade by request.