Generally speaking, it is more likely that a child will produce a low estimate of ability on a test than that an overestimate of ability will be generated. The CogAT is a reasonable screener for high ability, and usually does a good job of not creating false positives, but is known to miss some very high ability students (false negatives), especially those with highly divergent thinking styles.
As to whether the school is looking for high achievers...that is another question altogether, and best answered by looking at what GT programming consists of in your district. Is it faster, deeper, more creative, or just more volume of the same lower-level work, with extra executive functioning and output demands?
