Earlier quoted context omitted.
Only when they are disseminating in an academic venue. Most non-university research dissemination happens outside of academic venues. And even then usually only because it’s expected, not because it was actually useful. (And no, it’s not because academics are more ethical about acknowledging the shoulders they stand on. Academics rarely cite the chip designs, software libraries, lab instruments, instruction documents…
I don't find this true at all in my experience, you and I apparently have very different perspectives when it comes to the research we consume. The things which you say are not cited absolutely should be, and if they are not that's a problem. In the papers I read, I'm often encountering citations to specific versions of libraries, specific industry created operating systems, industry created programming languages of…
That said, even if we accept the general premise of your post, which I don’t, you’re still drawing the wrong conclusion.
To wit: citing something does not imply that the cited thing is “foundational” to the work from which it is cited. One can cite work for any number of reasons. (Admittedly, citation behavior did change with the rise of bibliomaniacs, but of course that further bolsters my overall point, so I’m not sure the daylight on this point does you any favors.)
You identified some counter-examples that miss the point because they’re unrepresentative, unresponsive, and irrelevant.
Unrepresentative because we are discussing literature in aggregate and this behavior is common.
Unresponsive because, in aggregate, inessential academic writing is systematically over-cited in academic writing and essential inputs of other types are systematically under-cited in academic writing. This is true of all academic writing; it’s a bias of the medium and of the medium’s standard bearers.
And irrelevant because there is nothing a priori or essentially nefarious about the above, on its own!
Academics beat ideas and lines of inquiry deep into the ground. Crucially, they do so by pumping out ridiculous quantities of PDFs. For every little variation there is a paper. Outside of academia this isn’t done. Eg: you cite Package X, great! But do you cite the 17 different PRs most relevant to your work, many of which are at least a papers worth or work? No. That’s culturally off. But for the corresponding thinly sliced papers that’s what you have to do.
Conclusion: academic work dominates the citation list because of publication and citation culture, not because academic work dominates the set of enabling contributions.
I do trust that you genuinely do experience the world as you describe here, but I think you’re a fish in water and that Upton Sinclair quote about paychecks comes to mind.