FWIW, I have yet to meet a gifted kid IRL for whom the CogAT was accurate. I know a few kids who did not score highly enough on the CogAT to qualify for a gifted program (in our district, 98th percentile), but private one-on-one IQ testing showed otherwise.
Accordingly, I think it's important to recognize that not testing highly on the CogAT does not mean a kid isn't gifted, and if CogAT is the only method of initial identification, a significant number of kids may be missed. I assume my view may be particularly applicable to students with twice-exceptionalities (using that term loosely), especially relative processing speed issues.
Was this using a timed version?
