fil et al. i will embed my comments as well

i said: "i agree that the boundaries between the inside functioning of and the human interface with an artifact are not always sharp.  in the design of scissors you deal with both, but in the design of a computer there is a clear division of labor.  interface designers do not need to know how bits are stored and changed, and the designers of a microchip have very little interest or place for how the icons look.  my suggestion is that human-centered designers declare an area of competence as their own and become so proficient as to be indispensible in the design of technology generally."

you said: I can see that it makes pragmatic sense to do this, given how "things are" these days.
i would argue things have always been that way.  designers never start from scratch or without the competencies of others.  after all very few designers start with designing molecules.  for example, in designing a pair of scissors, designers may want to take for granted how the steel is hardened.  architects hire statisticians to calculate the ability of a beam to hold the load of a building and ordinary occupants of buildings have no clue what is behind the walls - until it malfunctions.
you continue: I would just urge that we do (not) use of the boundaries implied by the creation of those competencies to partition up aspects of actual products and services that are artifacts of design processes.  The latter of which I really strongly believe is a "bad thing."
i'd say:  as long as steel manufacturers can advise the scissor designer that the material selected is inadequate or could be better, it cannot possibly be a bad thing.  similarly, as long as an architect hires a competent heating systems developer, he or she will get more than the architect can do individually.  ultimately this all is based on trusting the competencies of others.
you continue: This happens in engineering and outside it too.  An engineering example: a modern car is very clearly a "mechatronic device" in that it is really a complete blend of electronic, electric, and mechanical elements.  However, cars are still developed as if the mechanical and electrical/electronic systems were entirely separate things.  This artificial separation really hurts car development.
i'd say, it can indeed influence the development of a technology, but this is inevitable when designers of different kinds need to work together.  coordination can be achieved only in communication, using language augmented with graphics and models.  there is cooperative software available to facilitate that cooperation beyond the ability of individual designers -  to relate the topic back to jean schneider's articulation - without which truly complex computer systems can no longer be developed.  the result reflects the manner of cooperation.
you continue: My concern is that promoting a distinction between human-centred and technically-centred designers creates a similar separation that will be similarly disadvantageous.
i should ask you: disadvantageous  for whom?  surely you can't expect perfection when artifacts are complex in the sense of requiring the cooperation of many.
i suggested that: if designers know a little bit of everything and nothing deeper than what other disciplines have to offer, there is no special competence that designers can do research in, develop methods for, and offer to their clients, then they can easily be replaced by those who know can offer slightly more bits of knowledge of everything.
you replied: The special competence that a generalist has is in looking outside a given specialization.  I, for example, have a certain familiarity with formal logic that virtually NONE of my researcher colleagues have.  While it's a very tough row to hoe, I have been working on formal models of designerly activities that are descriptive (not prescriptive) and allow a kind of structured reasoning that is just not available elsewise.  Whether there's any /benefit/ to my work for practitioners or the world at large remains to be seen.  Breadth at the expense of depth can lead to innovation.  All I'm saying is that generalists see things differently, and we need both specialists and generalists to "see it all."
your personal competence notwithstanding, i was talking of professional design, of the content of focused design education, the discourse they need to develop to coordinate their work among themselves, how they as members of the community of designers present themselves to their other stakeholders.  i don't read you as advocating that familiarity with formal logical models is essential for designers to know, or that expertise in formal logic is the mark of a generalist.  i would say this exclusive expertise can often get in the way of solving social problems.  your expertise is an individual competencies, not shared by all designers, as you said.  as i argued, the distinction between the internal (mechanical and material) functioning of an artifact and the human interface with it, is fluid. nevertheless, the distinction helps focusing designers on what they are good at, need to have mastered, what others who know a little bit of everything cannot do as competently as designers can.
klaus