There has been an unsustainable inflation of academic research on the last 150 years or so after governments decided to formalize research. But the thing about unsustainable stuff is that they always end.
The institutions that teach researchers also doing the majority of it necessarily turn into a Ponzi scheme.
After thinking about it, I came up with a straightforward solution (at least in STEM): offer more than one type of of doctoral degree. Every program will have at least two doctoral programs: a Doctor of Philosophy, and a Doctor of Science/Engineering/Mathematics/etc.
The Doctor of Philosophy (PhD) program is academia at its core, where the students in this doctoral program are explicitly seeking an academic teaching or research position as their career path. The coursework and educational activities are explicitly aligned for this area.
The Doctor of Science/Engineering/Mathematics is focused on creating a top-of-the-line researcher intended for industry or an FFRDC. Those students receive a different type of education which explicitly gives them the deeper research skills and connections needed to become an accomplished industry researcher.
The two programs are equally rigorous but have different end goals in mind. This specialization is overdue, and most departments already have a fuzzy line separating the "academics" from the "practitioners."
Anyone who says otherwise is just either uninformed or selling a dream.
Your observation that the doctoral degree system has always been that way is precisely my point: the world has changed, and new forms of training are needed to complement the paper-publishers. The PhD system is broken in part because it's catering to multiple audiences when it should regain its focus on its core mission. That being said, many people want to do research but don't want to work in academia; in fact, I think their numbers are far greater than the academia-oriented. My idea caters to those people, and I think all parties (students, schools, industry, government, the general public) will benefit in this arrangement with almost no drawbacks.
From a degree-focused perspective, it's somewhat unusual that U.S. universities almost exclusively assign PhDs, save for the professional degrees (e.g., MD, PharmD, JD). Multiple types of bachelors and masters degrees exist, and those degrees are certainly differentiated from one another. In some European countries, the ScD is a terminal degree higher than a PhD.
I follow a dozen YouTubers doing extremely niche, cutting edge, science.
It is progressing beyond 'backyard science'.
Evidence of something that's been impactful?