Just to follow up on the first issue: my laptop has a touch screen. The framework I use (Love2D, which is built on SDL2) can detect this and receives multiple simultaneous touches.
In Harmopark the touchscreen is detected, but multi-touch doesn't work. Therefore I had to use touchpad.
In Harmopark the touchscreen is detected, but multi-touch doesn't work. Therefore I had to use touchpad.
Auto-voicing is complicated and it's very culture-specific. One good implementation is this: http://www.sealab2021.tv/etcetera/choir/