There's some innovation but there are also aggressive patent holders who are holding things back.
For example, Matius was stopping one-handed keyboards for a while.
http://blog.monstuff.com/archives/000021.html
There were things like the Frogpad or the HandyKey Twiddler
It is also, like many things in hardware, a software problem. The purpose of a keyboard is to interact with software. And that software is biased towards certain layouts. It is hard to change the layout too much without quickly becoming a niche product. Add to that the barrier to entry to making a hardware product is still quite high, and relatively costly.
I tried to see if I could make a keyboard with slightly larger keys for my upcoming live music performances and I was thoroughly disappointed:
https://old.reddit.com/r/MechanicalKeyboards/comments/9g0e24...
The DataHand was a new interesting “keyboard” design in the 1990’s, about 25 years ago.
https://en.m.wikipedia.org/wiki/DataHand
Btw, Xahlee has a great article on keyboard switches:
http://xahlee.info/kbd/keyboard_switch_mechanisms.html