Well, he's part wrong and part right as I see it. I don't expect Colemak to become common anytime soon, either. But let's pick at some of his statements a bit:

- He doesn't define how he assumes Colemak's superiority. I haven't seen any serious claims that Colemak beats Dvorak significantly when it comes to speed or ergonomy, only that it's easier to learn for the many QWERTY users around (and that a few other things work easier - like shortcuts and l/s etc.). Minor nitpick. The main error he makes at that point is to believe there are a "gazillion" Dvorak studies. There aren't. And it's not easy to make a decisive and rigid study, as it takes money and scientific skill to do.

- Each key on a modern board may take relatively little force to depress, but people still develop RSI. And people report being helped by more ergonomic boards and layouts. This rather defeats his theoretical argument that keyboards are lighter today therefore optimizing shouldn't be necessary.

- QWERTY boards are mass-produced, sure. But more and more of them are possible to remap physically now it seems. My kids will grow up seeing both kinds of layouts. That's nice. If anything akin to the Optimus becomes common at some point, it'll be even easier. Personally, I think that OLED technology will be absolutely huge eventually and we'll end up using them as living-room tapestries with the ability to light up the room in mood colours and show pictures that change over time and who knows what else, but that's going to take a while obviously. At some point though, I think OLED will be cheap and common since the principles of that technology should allow for it (cheap constituents, the possibility of printing OLED on materials with rapid printers etc etc). But I digress.

- Software remapping is already happening in at least two layers on his QWERTY machine (in Windows, there's the scancode remap in the registry, the VK remap and the final glyph send - maybe I've even forgotten something), so that argument is absolute poppycock. No users that I know of have reported slower computers from changing layouts. Although... I do have an absolutely bloated .klc file that might make an old 386 machine sweat... haven't tried it, but it could happen when remapping is taken to extremes I suppose. Bah.

- Employer-owned machines, sure. I haven't had any trouble at work though, and I work at a hospital with an IT department that is and has to be strict since people's lives depend on the hospital net. I'm not allowed to install anything on the hospital-net Windows machines, but I don't have to as I can simply run the portable and stealthy PKL on those. On the Linux machines and the University-net Win machines I have no trouble whatsoever. I think he overplays that argument.

- Indeed, people are ingrained and lazy and don't know their own best and it isn't going to change. So I don't foresee Colemak spreading like wildfire either. But I think it will become a standard because it has a few tricks going on Dvorak. As the various *nix distros will have it, the other systems will eventually come after. It'll take a while, but at some point I think it'll be available on most machines just like Dvorak is. Or the Serbian, Macedonian, various Sápmi and Gaelic layouts just to name a few. Lots of small groups get their own standard layout these days: Contrary to what your debatant seems to think it's very simple to implement on most platforms and there are enough people who care.

All in all, why bother so much about the "mass market future" of Colemak? As long as it's easy enough for each user to choose it, it's okay. It's not like it's Esperanto and you need someone else to speak it with!

One more thing: Saying that "the Dvorak software never became popular" is a bit odd. Firstly, no special software is required since the Dvorak is a standard on virtually all computers now - you just choose the Dvorak layout and type away. What software, did you say? Secondly, when this happened the Dvorak layout actually did become a lot more popular than before from what I've heard. Again, it's not been a sweeping wildfire, but the Dvorak has gained in popularity after it became readily available. And from what I gather, getting a physical Dvorak board isn't hard either.

I think there is a tendency for people to assume that new things either have to completely replace the entrenched standards or they are 'failures'.

I use a Mac, so I see this from time to time -- the old 'Macs may be pretty neat, but they're not going to replace Windows any time soon' argument. Sure, that may be the case... if your sample size is the entire computer industry. Macs will never replace Windows machines, if only because Apple only makes computers for consumers and digital professionals. But if you limit your sample size a bit, then things get interesting. Students, for example, buy Macs more often than five percent of the time. Limit your sample even more, and it can become really pronounced -- my parents and stepmother own four Macs and zero PCs between them. I own two Macs and one PC that I only occasionally use for gaming.

So will Colemak replace QWERTY? No. It can't, just because the 'target market' is the subset of users that will switch keyboard layouts. But with a smaller sample size, that changes. I use Colemak an order of magnitude more than I use QWERTY, and I imagine that pretty much everyone else on this forum is the same.

But does that make Colemak a failure? Does it have to replace QWERTY to 'win'? Is every QWERTY user still out there some kind of black mark against us? Who except our detractors ever says that the stated goal of Colemak is to become "common"?

In my life, Colemak is what I use for everything other than gaming and logging in to the computers at work. How does the fact that whatever millions of other people use QWERTY change that fact?