Easy editing with auto-detection of different voices
It would be great if the software could auto-detect different voices. Then the software could colour the waveform differently for each voice, so that it's easy to see at a glance when you're editing using the waveform in the timeline. There are various ways you could differentiate the voice displays; it could be as simple as having different colours, or perhaps darker or lighter tints of the same colour if it's preferable not to chance the clip colour.
To take it a step further, the software could auto tag the clip and give the user a prompt for "who is this person talking?", so that you enter the name once and then the clips and sub-clips are auto-tagged with the person.
If you have a large complex project with lots of different interviews, it could be very handy to be able to search for a specific person and not to have to manually tag everything.
and finally to take it even further - there could be an auto-edit function that would make cuts before and after the person talks. naturally the editor would need to clean it up and fine-tune the edit, but this could speed up the process of making an initial rough cut edit, for example by cutting out the interviewer's voice asking the questions in an interview (when they are off-camera and we don't want to include them in the video).
Recognising different voices is not hard to do for software like Izotope RX or applications that do automatic speech transcription, so it should be quite possible to include this kind of functionality in terms of automation and machine learning.