An update to this... I still did not find a solution for this, but found a workaround:
For each TTS.Speak Action currently in my commands, I added a previous TTS.Speak as follows. The result is that the first TTS action is executed but due to this issue with my computer/TV configuration, it will not actually output audio, but will turn the HDMI audio output on so the second TTS.Speak can be output correctly.
Not a elegant solution, I know, but works.
Notice I have changed the voice for this specific test. Check if you have it and change accordingly.
<?xml version="1.0" encoding="utf-16"?>
<!--VoxCommando 2.1.4.2-->
<command id="641" name="TTS.Speak test" enabled="true" alwaysOn="False" confirm="False" requiredConfidence="0" loop="False" loopDelay="0" loopMax="0" description="">
<action>
<cmdType>TTS.Speak</cmdType>
<params>
<param>o{CR}</param>
</params>
<cmdRepeat>1</cmdRepeat>
</action>
<action>
<cmdType>TTS.SetVoiceName</cmdType>
<params>
<param>Microsoft Hazel Desktop - English (Great Britain)</param>
</params>
<cmdRepeat>1</cmdRepeat>
</action>
<action>
<cmdType>TTS.Speak</cmdType>
<params>
<param>end I{CR} think to my, sel, aeaea aelf{CR} What ah, wuo wonder, faul: worl oh oh owold</param>
</params>
<cmdRepeat>1</cmdRepeat>
</action>
<phrase>Sing something good please</phrase>
</command>