Hey I actually send him the one month full-time prototype as a rough estimate :)
And I know that it is not that good of an idea, that is why I ask for time-based estimate, so that I would be able to send him to some od my freelancing friends.
The thing I am mostly affraid that solving combination of "voice-command"+"offline"+"loud kitchen" might prove more costly than anticipated.
But if you really need to go forward I don't see it extremely difficult...
Probably there are a lot of libraries you can use to manage youtube video on Android, as well with vocal command...
I will focus only in an American base version, it is easier and I don't see much market outside US.
I would say one whole months for a little prototype, working close to full time...