One modus operandi of such companies is to amass a lot of patentable ideas as trade secrets, which subsequently can be rolled out as sequential patents. This effectively extends their monopoly on the products as a whole, despite patents expiring that protect the initial innovations. They can always roll something out sooner if a competitor might be approaching their IP moat.
One of my friends works on an e-ink product. My understanding is that one of the major trade secrets are the e-ink waveforms (the sequence of voltages used to print and erase content on the display). They are shared under NDA and baked into product firmware. Apparently the open-source versions are significantly worse.
Hearsay, but I've read in the past that these were tailored specifically to the display, so difficult to actually collaborate since each display is a little different.
They only go out in devices after they've been patented.
Think of 3 years of research. Rather than it leading to three years of progressively better consumer devices, you instead patent the first idea and use it to go to market, then sit on the rest. After X amount of time (where X < 20 years), you patent the next winning idea, and go to market with it. Etc.
This presents a challenge to would be competitors; to go to market, you have to leapfrog the existing technology and patent (with as broad a language as patents tend to have), hope the incumbent doesn't have something that would immediately deprecate your product (or at least, relies itself on something you patented along the way), and then overcome the incumbent's existing advantage in in the market. And also be prepared for a legal fight, since almost assuredly one of you is going to accuse the other of infringing a patent.
If you want an e-ink device using 20 year old technology you can get a used one for cheaper than any would be competitor can produce one.
If you want one using anything developed in the past 5 years...oh, look, that's why eink still has a research division; they've been slowly improving things (not as fast as actual competition would cause, but enough to make it hard for someone to just leapfrog them using seed capital) and so would-be competitors are now running up against patents with 15+ years left.