Did you get that "plumbing" was just, like, an analogy? The article wasn't really about plumbing.
Computer science has always had this tension between being an actual "hard" science and an applied engineering discipline. A lot of otherwise intelligent people pretend to understand that distinction when they really don't and it leads to no end of problems.
I actually would go much further: I would say that the tension is between an applied engineering discipline and a vocational degree ala, well, plumbing.
I don't think there's much question that fields like quantum computing belong more in the "applied engineering" category (or even the "hard science" category). There's much more question about whether professional software development has any significant relation to the field of Computer Science other than the bare fundamentals.
My hunch in the future is that we'll see the field bifurcate into two very different professions, ala EE and electricians: "Computer Engineering" (or whatever it's called) will become a much more limited field that is overseen by the NSPE and has a licensure system, and "Software Development" will become a vocational degree (if that, even) and represent the vast majority of software development jobs we see today.
That obviously has problems, but I see it as an inevitable compromise that might eventually help resolve questions like, "this set of customer personal data was leaked because of a known but unpatched vulnerability, who signed off on this software release?" and also relives a lot of the pressure on regular software developers to understand these fundamental CS concepts that don't come up much in practice.
The conclusions he draws are questionable, too. Do you need AI powered search? Well, what if it's an elasticsearch plugin, takes 30 minutes to install, and increases your conversion rate by 5%? Should your engineers be required to know SQL and learn Redis on the job? Well, yes, they should.