I see your points....but I still think it could be a viable idea inside of Sonar. I was describing the basic idea which would need to be honed. I started thinking this way after I bought my last car and it has a lot of voice recognition in it...and by using siri.
There are ways to get around words spoken in the lyrics that might trick the system. This might involve saying " sonar record" or "sonar stop" or "sonar arm" ...you're not likely to have those in your lyrics :)
The implications are actually better than touch screen in my thinking...you probably can't use it for automation or fader moves, but tracks could be selected ex. " sonar track 3 arm" "sonar track three record", "sonar track 3 rewind to 32"
This really isn't as far fetched as it may seem and think of the handicapped people who might be able to make use of the technology in addition to making a tracking session a piece of cake for a guitar player who is wearing his axe.
Another audio line would probably be needed seperate from the main tracks intended only for voice/speech recognition.