They mean a platform for developing for eink, not a development platform using eink.

Lots of people want to make projects with epaper displays but it’s not particularly easy or cheap to source the panels and hook them up to a Linux system. Reusing these ereaders makes it cheap and easy.

The main issue with this is that a Kindle is not a typical Linux environment; they use heavily patched kernels that range from old to hideously old depending on the age of the device, have a severely stripped down userspace similar to most other embedded devices and it's a pain to set up the development tools needed to compile software against the various libc/kernel versions used across the Kindle family.

That's fine if you're using the built-in kernel as a souped-up display driver. All you need is a way to make pixels happen, the rest can happen off-device.

Exactly I'm in a process of turning one Kindle into a openhab panel.

I failed a bit during the soldering part (the pads got unglued, so now I need to push hard a conductive cable, until I get wifi and ssh up and working).

This post will greatly help me because I mostly need that: wifi and ssh working (and then make the pixels happen :), but I saw a solution for that on mobileread forums).

I suppose "development target" might be more apt. The whole purpose is to be able to write software for other people with minimal messing about with the kindle to see that it works.

So, for example, if you had an inclination to write various pieces of software that turn a kindle into a wall mounted device, this would be the meat of your pipeline. I could write a home automation panel, then later, a weather display, etc. The prior work means I can push those things and make sure they function well before releasing them to the public.

It's needed because it wouldn't be easy to replicate the Kindle environment on a normal PC. It uses old custom kernels, a very limited subset of userspace utilities, has an unusual display, etc.