OK I figured it out, I'll put it here to help someone in the future. The key point is the pitch scale change changes the currently PLAYING sound, so it must be used after the play command.
I now have this:
voice_set_pitch_scale(0,1.2) the first argument (which I have as
0) is the index of the sound that's playing, so in my case, the first (and only) sound.
The second argument, which I'm testing with
1.2 is the scale change, so 20% higher in this case.
I hope that helps someone in the future.