My experience with the CogAT is generally not positive. At one point there was a >70 point split between kindergarten CogAT verbal (SAS in the 90s) and VCI on the WISC-IV (in the extended norms range). It's not necessarily a good instrument for identifying every child who may be gifted, but it does seem to work relatively well for discriminating MG from average most of the time.

"The thing that doesn't fit is the most interesting."
-Richard Feynman