Are nonfiction writers, that avoid the use of footnotes/endnotes, respected by the academic community?

I don't see an issue in many (most?) cases. Especially if they are writing for a popular rather than academic audience. The question is the tradeoff between readability and the need for specificity in such works. Let me give an example.

While I'm a mathematician who taught CS until retirement, I read quite a lot of books on anthropology and archaeology. I don't read scientific papers, but try to follow advances as reported in the press as they occur.

There are two kinds of books. One in which an author will give a narrative covering what is known about a topic (say early humans in what became North America) without inline citations. These can be very readable, I find. There is no danger of assuming plagiarism since the author(s) give the further reading section which lists the source. Most of the ideas presented are "in their own words" rather than copied from sources, and most represent a scientific consensus along with some information about where the consensus breaks down.

Other authors, writing on the same topic use a large number of inline citations, with footnotes or (usually) endnotes. I find these to be much harder to follow and it is often necessary to go to the endnote as more is presented there than just the source, such as a few additional ideas. They may be a bit more precise, but for the audience, not essentially so.

So, in some sense, your question is, whether popularizers of science are respected or not. Some are, others not, of course. Great story tellers are valued, I guess. But it is a choice that an author makes. The ones I read are careful not to plagiarize, but are a bit less specific about which ideas come from which source. However, once a scientific consensus is achieved, even with some dissent, it seems less vital for these types of work. The "further reading" gives you a sense about how the consensus arose.

But if you are writing at the edge of the known world, the rules about citation are much more vital, since a consensus has not yet emerged.

Your question does not address an attribute of a person (in this case: J. Diamond), but of a publication - or to be more precise, of the social system (e.g. system of scientific publications, system of popular books).

If X writes a paper for a scholarly journal, X needs to use footnotes or in-text citations because the scientific system requires the ability to trace the origins behind a statement. Otherwise, the system will reject the paper (through peer-reviewers and editors).

But if X writes for a popular press, such references are not required; editors may even reject book proposals if they seem 'too scientific'.

Thus, Jared Diamond uses footnotes or in-text references when he writes for an academic journal (such as this paper for the journal Biological Conservation). And Google Scholar tells us that this peer-reviewed paper has been cited more than 2.000 times in scholarly works. There therefore seems to be no lack of respect for Diamond's scientific outputs within the scientific system.