I have been praying for a GUI library that at least supports a skeuomorphic style for UI elements; this, however, requires some creative designers without preconceptions ready to take on a challenge. The challenge is to create a forward-looking skeuomorphic design with subdivisions adapted to different devices and workflows (separated e.g. by screen size and by expected expertise of the user).
This is hard to do when most designers are stuck in the non sequitur that skeuomorphism means a dated 90s feel (a baffling lack of imagination for a seemingly creative profession) or that a single design instance should support all devices (I will not use a word processor on my phone unless absolutely necessary, and I sure as hell don't want to touch-click gigantic rectangles on my widescreen Microsoft).
And of course, as you say, it is important to read the rich HCI research conducted over many decades since the 60s, and not dismiss it as "old".
Well, I'll say that I won't just support skeumorphic design; it will be the only option. :)
The rest of your post lays out my plan perfectly.
Yes, it will be a challenge, but I think you are absolutely right that UI's need to be designed for their audience, like blog posts and other writing.
Now the question is: if you need to use such a library in a commercial context, would you happily pay for it? I can't justify the time unless I could make money.
This is hard to do when most designers are stuck in the non sequitur that skeuomorphism means a dated 90s feel (a baffling lack of imagination for a seemingly creative profession) or that a single design instance should support all devices (I will not use a word processor on my phone unless absolutely necessary, and I sure as hell don't want to touch-click gigantic rectangles on my widescreen Microsoft).
And of course, as you say, it is important to read the rich HCI research conducted over many decades since the 60s, and not dismiss it as "old".