When the "Say Something with Text-To-Speech" action has its "Voice" option set to a voice that is not available on the system (E.G. the literal text "this voice does not exist"), the TTS engine will not stop speaking when "Stop Commands" (by shortcut, UI button, or action) or the "Stop Text-To-Speech" action are used.
The "Wait until speech completes before continuing command" option does not make a difference.
This occurs at least as far back as v1.7.3
EDIT: Fixed in v1.8.3.22