Regardless of the recognition setting, the speech recognition engine will wait until you stop speaking before attempting to recognize another command.
E.G. if you have a "Restricted Continuous Speech" command named "my override" and another command named "my command", and you speak "my override my command", only "my override" would be recognized and executed. You'd need to speak "my override...my command" instead.
The "Override listening if my spoken command begins with" option allows you to speak in a single sentence.