Hiya Pinkyache! Long time no see, right? Good to see you.

I don't have any direct responses to your post. Just saying hi, for now.  ／(⁎≧∇≦)＼

It's possible that learning Colemak - or any other decent layout - would improve your technique, such that you could then apply those lessons to Qwerty touch-typing as well? 

Even if that were the case, I can't see how that would be better than just learning touch-typing in Qwerty direct, or better yet, learning to touch-type in Colemak and then ditching Qwerty entirely. The backward step to Qwerty after become skilled in something much better is not going to come easy.

As an aside, I think quite often these sorts of articles are designed to be comforting and reassuring to Qwerty users as the writers don't want to alienate the majority of their readership, most of whom have no desire or interest in switching.

That kind of claims makes me laugh, they're just pulled out of people's anuses. Where's that number from? It smells of offal.

Typically, someone did an ill-advised calculation, or performed a poorly-designed experiment then just gobbled up the results thereof. Highly untrustworthy.

That said, it does seem that better layouts don't offer much of a raw speed advantage. The fastest typists say that even at 220+ WPM they don't feel that physical factors are the most speed-limiting. Rather, it's the mental task that keeps them from even higher speeds. And that task isn't highly dependent on layout efficiency I think.

Thanks for the reference! I see that in this particular case my critique was too harshly formulated: EngineerGuy did look up rigorous attempts at measuring real layout efficiency.

Unfortunately, the issue still stands. Even the best studies to date have done little to measure how people actually learn a layout properly, to the best of my knowledge. Such an actually interesting study would be very costly and time-consuming to conduct. I think it's fair to say that we may be able to measure relative quality between layouts, hopefully fairly well, but not really numerically.

What is probably true though, is that if speed is your only endpoint then better layouts likely won't be much more than 5–10% more efficient after all. As mentioned, even top typists do not feel that speed limited by different layouts. Jashe famously learnt QWERTY, Dvorak and Colemak to ≈200 WPM for a MonkeyType LayoutFluid challenge, and reported that Colemak feels best at that speed followed by Dvorak (which had a little too much alternation for such a high speed according to him). And while it's easily seen that the top typists using QWERTY is more a matter of habit and experience with it than anything else, this is still the situation. It's quite easy to brush off any dreamer's claim that their new layout is n% faster or something like that.

Indeed. And another bump in the road for the debate, is that you and I at least may not be able to discern the differences very well at our speeds. Hell, I type slowly enough that I'd probably find even Workman comfy enough. I try to keep an ear to the ground and listen to what the best typists say about layout quality, so I can join in when people discuss it – but I may not be fast enough myself to really feel any of the differences very well.

I still like my layout choice a lot of course. It's soooo nice. Especially when adding in the extra bells and whistles of the Big Bag.

Great reading and you are right Colemak QWERTY felt quite unique or different on fingers. But my point of view is that learning the QWERTY is good thing but taking it in routine rather then touch keyboards is not a good option. I may be wrong but it's just my opinion.