Earlier quoted context omitted.
Is there actually any evidence ergonomic keyboards reduce risk of RSI? Last time I looked into it it seemed like marketing fluff with no actual studies that backed it up.
RSI is a complex thing. Some people have pain that is fixed by uniquely shaped keyboards, others need to just switch ctrl and capslock, get rid of the numpad or whatever. I don't think there is such a thing as a keyboard that cures any kind of RSI. But I do think something super flexible like the Kinesis Freestyle comes close.
Like Jeff, I was not content with any of the available keyboards. I have two Kinesis Freestyles (the old one and the new one), and they are both great. But the halves only separate by 20", which is not enough to mount them on armrests.
So, I cut the separator cord one night, only to find that it contained twenty fine, individually shielded wires. At length, I reconnected them with an extension piece (three CAT-5 cables), first with breadboard (proof of concept), and eventually with soldering, heatshrink wrap, etc. I am NOT a hardware guy, and this took forever. But it was worth it.
Freestyle nitpicks: (1) The hardwired keys along the left (C-x, C-C, C-v, etc) are pointless and easy to hit by accident, yet hard to find on purpose. (2) "B" should be on both sides. (3) It would be nice if the two spacebars were distinguishable. I'd map one of them to Ctrl.
Mapping caps lock to ctrl is invaluable. (Except when you're on someone else's machine and you're constantly WRITING IN CAPS.)