So what's really that different to now from then?
So, this will 'only interest a small percentage of the schoolchildren' and 'therefore isn't really viable' is only a challenge in these modern times?
Back in the early 80s home computers were treated by most of those who chose to take any interest as a way of playing games, but everything else the computer could do was readily accessible; nothing was hidden from you if you chose to look any deeper, and some of those who did play games went exploring, usually to try and 'write their own games'. Out of a school of 300, I'd say only 30 or so kids took much interest at computers at school, and maybe 10 for anything other than playing games.
Now, most computers are treated as devices by to play games or "do something on the net", but everything about the machine is so abstracted that if you do want to look any deeper, you're likely to be out of luck. That same school of 300 would likely have 290 who use computers, but almost zero who have any way of delving beyond that tightly-controlled veneer.
Sure, the Pi and its ilk would probably not manage to get more than 10 or 20 kids interested in the guts of computers and computing, but even getting it back to the percentage from the early 80s would be a success; Those skills I learned had uses FAR beyond simply programming, so not even providing the tools to offer the chance to find out is a tragedy in itself.
Microsoft said “Where do you want to go today?”
In my opinion, that little flashing "K" in the corner of the screen offered far more potential.