Ok, coding != software development, ...

But considering that it's usual to expect a new software developer to take 4-8 weeks to start being productive, I somehow don't see tickets being distributed via Amazon Turk to some Indian coders, ....

You forgot <<

You do realize, that all these fields nowadays strongly depend upon Computer Science for their research and data processing.

Computer Science is similar to Mathematics, that it's an universal auxiliary science, that is usually used in other fields.

Notice that CS majors usually work in an industry, and only very seldom (if they don't go academic) it's computers/IT.
IT as an "industry" is usually only a cover for we are doing "general work" for other industries.

Absolutely concur, but than, it might be my personal bias, because the trivial stuff like programming I self educated myself as a teenager, Which ended in a situation where I was taking an "introduction to programming" class at one university (in Modula 2), while being a tutor supporting students in an "introduction to programming" class at the other university in city (that one was in C++).

You can envision how useful that "introduction to programming" (in Modula 2) was. Although I have to admit I was forced to learn Modula 2, what a useful skill. And I had some fascinating insights (the prof had the great idea to let peer-review be part of the grade, hint: global variables are your friend, that magic local variables and argument thingy are bad for your grade).

Well, then you work in the wrong industry.

There are places (e.g. aircraft industry, OTOH, writing control software for Navy ships capable of shooting down airliners, funny as it sounds does not require that level of correctness, sigh) that require formal verification for parts of the software, and then it's nice to know the limits of the methods you apply.

Funny that you mention it, they looked the same 2 decades ago ;)

logical thinking and worse, abstract thinking.

You are mistaking Computer Science for IT. It's like mistaking the ability to add and multiply for mathematics.

And how does that make Computer Science change? Just because our cars today move so much faster, have commonly air condition, and so on, does not mean that the physics has changed in the last century.

I think the great misunderstanding is that Computer Science is named as it is, so many people think that it's only about computers. And the second problem is that US institutions have been, at least at the B.Sc. level been very "practice" oriented, which sounds initially a good idea, but is actually a bad idea. Because learning a new programming language is a thing that takes a week, perhaps two. And after the first dozen (in random order for me: different Assembler languages, Basic, Pascal, Modula 2, Modula 3, C, C++, Objective-C, Prolog, Perl, PHP, Python, Haskell, Lisp, Javascript, Lua, Erlang, Fortran) it's rather routine.

The difference is like knowing SQL, relational algebra, and DBA skills. Of these relational algebra and DBA skills are valuable: relational algebra because it's the theoretical under pining of SQL, and DBA skills, because they are very hard to get (it's a chicken and egg problem, it's seldom that one gets a chance to enter the field, and it's not something that you can learn on your private laptop, only in practice with big iron). Learning SQL on the other hand is yet another language. Not even Turing complete.

In IT much has changed.

In CS not so much has changed. And the changes that we are seeing are less breakthroughs as such, it's just that the hardware has changed drastically, so things that were just unthinkable two decades ago, are trivial enough.

Just think, the MicroSD card in my mobile today, has a capacity that is a million times higher than the floppies that my first computer used.
The cpu in my mobile has more cpu cache than my first PC had memory.
The cpu in my laptop has slightly less cache than the size of the first PC harddiscs.

So, yes there are changes in computer science, but these are driven usually not by fundamental breakthroughs, it's the hardware. E.g. currently the next "new paradigm" is the observation that even SATA based SSDs have totally absolutely different performance characteristics to HDDs, and if you use high end pci-e based SSDs, it's even more extreme. That triggers research/development of new algorithms and designs, but it does not invalidate a real CS education.
The "storage hierarchy pyramid" you might have learned about just got a slightly different look, but the concepts as such still applies.
 

Well, considering that many "experts" consider HTML a programming language, hence writing HTML is coding, ...
sigh

Well, from my experience, they are masters at milking the situation. Including not having all the necessary parts despite a description of the situation when calling them, so that they need a second visit that you have the privilege to pay. ;)

You do realize that courts have ordered companies to hand over their private keys to the FBI in the past? Some companies like lavabit have closed down over that, others have surely complied without informing the public (not all CEOs/legal departments realize the relevance of this).

Well the government did it already, remember lavabit?

Another aspect other countries have courts that can issue legal orders too.

Winner how the FBI will it like if the PR China decides that Apple has to cooperate or they will be removed from the Chinese market.