I agree with a lot of what you write, and I disagree with other parts, but what strikes me most is that you touch upon a big issue in software development the past 10 years. From around 1995 onward, a large number of software people have invested in software architecture as a discipline. A discipline that had by the early 2000s had established clearly that architecture is not just a about defining technical structure/blueprints, but just as much about organisation (customer as well as developer side) and process (same).
It seems that the past 10–15 years more and more software engineers were hired and appointed, who were not trained and coached to see that. That’s part of why concepts like microservices (and before that SOA and before that component based development) get applied when not applicable.
Your article addresses part of that, and the comments confirm it. Begin agile is one thing, being at the front of new technology is another, but there is a reason we spent 1000s of man years defining what architecture is, why design is important and why we can’t do without proper testing. We should apply that collective experience, that knowledge base, and stop reinventing the wheel by making the same mistakes over and over again.