In my experience selenium cell meters that are not accurate demonstrate a response that is difficult to correct. For example in bright light it may be 1.5 stops off, but in dim light it may be only 1/3 stop off. Did you check the Sekonic across the whole range with your Gossen?