In addition to RealFred's points: the motivation to learn comes from some future perspective or an unmet need. A barrier to participation in IT is that kids now have few unmet technology needs. Can't be bothered even breaking wind; there's an app for that. Have a great idea for a game; organising or shooting? Robotics; cost versus benefit and industrial robotics already do it better and faster. All this before we get to the 457's and the patent trolls. Gone are the days when 10 Print "Hello World", 20 GOTO 10 was the leading edge. This was the mistake the Raspberry Pi team made. Teaching more kids programming / IT was not about accessibility, it was about needs.
To get kids doing IT you either make them do it (a need to pass a test) which will kill any future interest or you find a genuine unmet need and encourage the kids to solve it before every man and his dog writes an app and patents the idea or governments decide it is a threat to mass surveillance and control. To make it accessible to the many the need must be able to be constructed lego block like - ie at a high enough level of abstraction that you are back to teaching applications, configurations and themes not programming per se and it must be on a relevant device, their smart phone. The completion of the need must also convey a sense of ownership and empowerment therefore you are talking open source, possibly client and private server (role for the Raspberry Pi?) and definitely networking, encryption, certificates, revocation and stenography.
Perhaps coming up with engaging lesson plans and the research of unmet technology needs in teens is an area where the CSIRO could spend its time and money. Though I suspect that the organisation has become so politicised and politically correct after a decade of global warming theology that it would struggle with this task.