This is the absolute, most common, frequently found, frequently followed process for developing software. There is more "wiggle room" at each step, but it's the same exact process, whether it's one guy doing it and dreaming up the requirements or a team of thousands of developers.
For instance, the people that write papers on new datastructures, computer vision and other state-of-the-art developments. They're not working with 'requirements' other than 'it would be nice if you found a way to solve 'X', come back in three years'. (well, not quite like that, but closely).
That's basic research and a definite amount of artistry is involved at that level.
But after the paper is written up it has solidified and it becomes applicable science, and for the person calling the API that embodies the concept it has become 'mere' bricklaying.
I completely disagree.
The art for the person calling the API is often found in "building something people want." Think of the people who frequent this website. Many of them are writing lines of code to make something people want. Surely there is some art to that. There needs to be an idea, building something to show customers, and iteration on that idea taking customer response into account. The customers have input, but the programmer decides what to write.
Just as wealthy people once commissioned portraits and told the artist what to draw, programmers can be employed someplace where the question of what to program is decided by someone else. But there is nothing inherent in art prohibiting painters from deciding what to paint, or in programming prohibiting programmers deciding what to program.
The vision you seem to be espousing is someone tells the programmer exactly what to do, and the programmer does it exactly to that specification. Wiring a house and balancing a budget are solved problems. So you seem to be saying that programming is also a solved problem.
Which is so out of touch with reality I feel that, surely, I must be misunderstanding you. A computer program can do pretty much anything. Depending how you feel about the equivalence of Turing machines and human brains, you can argue that a computer program can potentially do anything a person can do.
With the geometric increase of computing capacity available to us, how is it even conceivable that "programming" is equivalent to balancing a budget or wiring a house? You seem to suffer from an incredible lack of imagination.
"However, I think that is a much better position to aspire to than self-aggrandizement and glorifying a skill that is really no different than any other skill that any suitably motivated 12 year can acquire with a couple library books and an internet connection."
That is the same as saying a 12 year old can write or play a musical instrument. Sure, they can spell words and make sentences, or know the correct fingering for certain notes. But the potential for improvement in those endeavors is also unbounded. The same is true of programming, in my opinion.