Right and wrong
Kris is right in his premise, that the Raspi is not by itself the saviour of British computing. But his arguments for that are aimed badly.
1) Yes, most household do have a computer. But they don't usually have one that available for enough time for development. Case in point - at home we have one PC shared between 4 people. That's simply not enough time for me to sit down and do dev work without someone else needing to use it. Also, PC's are not great for any sort of HW interaction.
2) The Pi is easy to tinker with. Yes, it is, considerable easier than a Windows PC. Arduino is also a good option here, but needs a separate PC for compilation - the Raspi can do everything on board. Or attach an Arduino board to the Raspi for the best of both worlds.
Note that the uUSB connector is not fragile - it's rated for many thousands of connections. Or you just turn off at the mains...I do both.
3) Linux is not easy. True. Easier than it was. But don't underestimate the knowledge absorption capabilities of the average 14 year old. If they learn Linux at that age there will be no stopping them.
4) Lots more bedroom programmers. Difficult one to assess. There may well be, but the arguments put forward are not related to the Raspi at all. Yes, there are many languages to choose from (but, TBH, just go with Python or C), and read up on the turorials from the MagPi or from the coursework being produced for the new curriculum.
5) Well, I think you are stating the obvious here. Of course, by itself the Raspi is not going to change to the world - it needs teaching material, good teachers, the right languages (easy - Scratch for the young ones, Python and C and perhaps Java for the older ones) - all those things. But then, so does any other teaching device.
But most importantly - teach the concepts and logical thought. And when only 5% of those you teach think this is great and show aptitude -' I'd like to do this for a living' - then you have a success.