I suspect touch-typing-optimised keyboard layours are not as relevant now as they would have been for the typewrite generation. My default home position has my

I suspect touch-typing-optimised keyboard layours are not as relevant now as they would have been for the typewrite generation. My default home position has my right hand in the conventional position, but my left hovering over the bottom left-hand corner of the screen. I spend far more time using those meta keys than typing. Ctrl+e and ctrl+y (vim scrolling keys) and alt+tab (window switching); ctrl+pageup/down (browser switching) and ctrl+a (gnu/screen escape key) than I do bulk entry of text. It also puts me closer to the escape key. In Windows I heavily use control characters based on win+r, win+e, and tab and shift-tab. I suspect the reason I have the right hand in the 'correct' position is because the dimple on the keys there allows me to consistently return to a known position in between occasional mouse usage. A small change to the keyboard layout (e.g. trying to use a UK English keyboard when I grew up with US English) is annoying not only because of the different positions of some control characters I use a lot, but also because it changes the feel of the keyboard and gives a different feel when I'm trying to return to a known position. I can't slide along the left shift there to navigate my left hand to the right position. I suspect I'd have an easier time learning a different combination of alphabetical characters than I'd have adjusting to changes in the shape of the keys. I've not seen a typing survey about 'optimal layout' that considers such things, and given how individual it is I'm not surprised. The only reliable way to get figures out would be to take different populations of people at birth and grow them into different layours. With the entrenchment of qwerty and the tie between applications and particular layours (e.g. navigation keys in vim, popular positioning cut/copy/paste hotkeys in qwerty and the important role those keys play in users moving away from the mouse and to keyboard control) I doubt such a study would be practical.