Everything needs to be separate so whoever is putting it together can have control over every set of vocals. So if you record your rap while listening to the song through headphones and get your girlfriend to do the same on a different track, you will have two files to send to the person putting the song together. Then they can control the volume for each person separately.
The same goes for everyone recording vocals. Record your vocals with no music on the track. Listen to the song through headphones when you are all recording. If the music is there it will get really messy and noisy when the song is put together. We need the best possible quality for this if it's going to be released to the world :yes:
I know there's no such thing as perfect but there's no harm in trying to achieve it either
EDIT: The demo from YouTube isn't working anymore. It says 'removed due to terms of use violation'