An asymptotic expansion of a confluent hypergeometric series is used to approximate the exact finite sample distribution function of a nonconsistent GCL structural variance estimator. A theoretical result is used to motivate the specification of a simple algorithm under which we may accept or reject the use of the asymptotic distribution function of the GCL estimator to approximate the exact distribution function.