Thanks for this interesting take of the age old argument!
I thought about this topic a lot and want to offer my 2 cents. I'd like to think of "the should UX and UI be separate argument" from another conceptual angle.
Why must we focus on UX+UI? Why not UX+research? or UX+Coding? or UX+UI+Coding? or any permutation one can think of?
Like a lot of things in life, the answer is always "it depends".
It depends on the company.
For example, Basecamp is known for asking their designers to be able to do UX+UI+ little bit of coding / prototyping, like the example you given for the tech companies.
It makes sense for them because this is what is needed for their company.
Likewise, using your surgeons analogy, I would say it depends on where the surgeon is. If you work as a medic or a UN doctor and you are in the field hospital with a patient in front of you with some medical condition that you are not specialized in and think you can't work on it. Guess what? Now you do.
I agree with your point that "my company needs you to do x, y, & z", if you can only do "x", you are probably not for us.
There is no Right or Wrong, only what is needed for the company.
A final point I want to make is that specialization is ever changing depend on the market condition and knowledge the market acquired. For example, 500 years ago where is no medical specialization because we didn't know what we don't know. It is when we discovered that the human body is too complicated for one person to know it all had specialization occur.
However, once we figured enough out, some knowledge will be consolidated to become general knowledge while other front tier knowledge will remain specialty of someone. Any general doctor today will be seen as a unicorn by someone from 500 years ago because they can out-knowledge any specialist at that time.
Perhaps your point is correct where we are at the stage where some skills has to be consolidated (UX+UI+protyping), but I'd like to think generalist and specialist can also co-exist. For example, anyone that is really good at VR/AR design right now will be the actual specialist for those domain. Hard to make the same argument 20 years later when every designer is expected to know the design patters for VR/AR.
