Correct me if I'm wrong but didn't Librettos skip from the Pentium MMX to the
Pentium M? So who enables such machines, Toshiba or the component suppliers?
Toshiba could only build such computers because other companies made suitable
components. Toshiba deserves credit for taking the opportunity but it's not like it
was all their own work or an idea that was entirely theirs.
If you go far enough back then companies like Grid and more broadly Sharp, Tandy
and Psion deserve some credit for the 'computer in your pocket' idea.
So why are netbooks (smaller than laptop PCs) back in the limelight?
Three things stand out to me:
1. Cheap, fast-enough, low power processors. A PIV netbook was never an option.

I've already gone through this, but having a part designed for rather than adapted
to the goal makes a difference.
2. Cheap displays; the panels used in many of the netbooks are cheap because they
are being used in other devices such as in-car entertainment, and digital photo frames.
If they weren't mass produced building a small netbook would be much more difficult
as small high res screens have been quite expensive until recently.
3. Flash prices fell of a cliff. 'Enough' flash is now affordable, and of course it's low
power as well. But in the context of a netbook flash beats an HD because - a) it is
smaller than standard 2.5" and 1.8" HDs, and even the smaller drives for modest of
amounts of storage and b) if you only need a few GB it will cost just a few dollars,
there is no just a few dollars HD, the '$5 HD' does not exist.
Most of the other components haven't changed as much in terms of technology or
cost. Batteries, back lights and chip sets are likely to improve in the future and
assist further reductions in cost and power usage.
I'm not disputing that small PCs existed before, but I do think it's clear that several
trends have come together to make them more available and probably successful
then they were before, or indeed could have been. That's the new bit.