Speech recognition combined with text expansion, or stenography can absolutely destroy the fastest typists.
The point here is whether typing speed is relevant. I gave an argument for why it is.
Now, whether one can achieve higher speeds with a stenograph a
different issue: it is about the best device to type fast, not whether it is relevant to do so. I do not dispute that using a stenograph can be faster than a traditional keyboard, but they also have many downsides: they require special hardware (a n-key rollover keyboard can be used, but that still is special as typical keyboards do not have this feature).
Speech recognition is more limited than a stenograph and it is very cumbersome. Either it is noisy and goes against privacy (which may or may not be required) or requires uncomfortable special hardware (a "stenomask"). In most cases where a keyboard is used, speech recognition can only complement but not replace the keyboard, due to the relatively high error rate of the former and because keyboard shortcuts are very cumbersome to enter (that is it, if they are allowed by the speech recognition system in the first place).
So the concussion is that stenographs and voice recognition systems are not an universal substitute for keyboards. Specifically, if you are not going to use a stenograph or voice recognition system, then the fact that they may be faster is a moot point.
Meanwhile, *typist* as an occupation is scarce nowadays, precisely because we have copypasta, speech recognition, OCR etc.
Indeed, meaning that almost everybody does his own typing himself, instead of having somebody else do that for him. That is an obstacle for specialization and a further reason of why stenographs are unlikely to become widespread (It pays better for a specialized typist to go though the hassle of learning to use a stenotype and purchase the required hardware than to somebody that just types as part of his activities).
Better yet, many members are typing on keyboards made in early 90's or 80's.
Also, durability or the possibility of repairs isn't about price per se; it has environmental and social consequences as well.
That is right, and those environmental and social consequences are worth examining. I can not comment much in that regard, but I can note that the environmental friendliness that a more durable keyboard may have is
countered or possibly reversed if its user constantly consumes more goods and services just for the sake of customization (like custom keycaps). Remember that the environmental cost is not only the materials that compose the final goods, but also intermediate consumption, packaging, transport, and the environmental consequences of the extraction and processing of the raw materials.
As for the social consequences, I am not sure of what you are talking about. Could you please elaborate?. Maybe you are referring to the common idea that consumption creates (more frequently, preserves) jobs; however, for that matter, if one doesn't spends money in an expensive keyboard, one is going to spend it in something else that also generates economical activity.
Developers do very little typing in actuality, esp. compared to interaction with the interface, and the dumb work is slowly being replaced by automation.
I agree. I do not work as a programmer, but I sometimes write computer programs. I notice that I spend more time only thinking than typing (and thinking), and bear in mind that I think while I type. I conjecture that this is the rule rather than the exception except for intellectualy trivial tasks
(see what I wrote above). If you
really type that much, to the point where typing speed is your limiting factor, then you must be writing something trivial in a very inefficient way (maybe manually unrolling loops without using copy and paste or kill and yank). Also see
http://xahlee.info/kbd/how_many_keystrokes_programers_type_a_day.html.
