Sorry that it is so difficult to search this forum, but there are a number of threads that address this issue from various perspectives.I have seen the comments on the decatxt thread which are very negative, the decatxt looks rather uncomfortable to use and I can understand why the average user probably wouldn't want a chording keyboard.
Many (98%) of the comments are negative, but it is always helpful to see past mistakes so that you do not go through the exercise of repeating them.
Did you explored http://chordite.com/ ?When I was only considering this project I tried to explore other already existing chording keyboards and the chordlite was one of my favourite looking ones, I am not sure about having 2 buttons per finger though but in the video it looks rather pleasant, it also has the advantage of leaving one hand free for mouse or pen usage.
It's a one hand keyboard but it's the best issue I know for that kind of stuff.
It keeps occuring to me that if you modify the datahand key design I've mimicked here:
http://geekhack.org/index.php?topic=41422.msg882543#msg882543
To remove the East and West keys, then you might have a perfect chording setup for one hand.
You would need a different pcb layout in order to pack the keys in closer, and more inline,
But you'd get 12 keys instead of 4. PM me if you're interested.
So a day of tinkering and I now have 1 'hands' worth of keys mounted on their 'plate' (made of molding clay) and wired up such that they can change the colour of an LED.
The Teensy is turning on one pin of the multi-coloured LED when it detects that one of the keys is pressed, it has no idea what chording or USB is yet though.Show Image(http://i.imgur.com/2FTKlNJ.jpg)Show Image(http://i.imgur.com/zQiYmQC.jpg)
So now 'just' a matter of doing this all again, building a grip (or more likely, temporarily using the play-doh prototype), figuring out the Teensy 3.0 USB libraries and then writing the firmware....
progress, I promise ;)
By searching US patents it quickly becomes apparent that an awful lot of people have worked on alternative keyboards. It's what they call a densely covered subject. There are dozens, hundreds really, of patents on every aspect. Of course every inventor is convinced his gizmo is the best and best possible. Oddly however the patents just lie there and expire fruitlessly. By the hundreds. So if you adopt a scholarly approach, rather than just leaping into tinkering, you will learn about a whole lot of ideas that aren't good enough in one way or another or, usually, in several. Now for me that was good, because I don't get off on tinkering for its own sake. I do like to learn stuff, even if it's just stuff to avoid. Long story short: learn from others' mistakes. Build yourself a Chordite. :-)
https://github.com/imarko/dhteensy
You should be able to jumpstart your S/W by starting with dhteensy...