Currently, if you do not tick the box to make the text to speech command wait until it has finished speaking, you can finish up with a situation where several phrases talk over each other.
Conversely, if you DO tick the box, the rest of the command doesn't get done until the TTS has finished teling you it is going to do it.
It would be helpful to have the option to put the TTS calls in a queue so that they are said one at a time, but the rest of the commands are executed promptly.