Style guides on structured/ordered bibliography?

Question

Edit: The best guideline I have found so far is the Chicago Manual of Style 16th ed, section 14.58 "DIVIDING A BIBLIOGRAPHY INTO SECTIONS" (I have boldfaced the part which best answers my question):

A bibliography may occasionally be divided into sections--but only if doing so would make the reader's job significantly easier. Where readers need to refer frequently from notes to bibliography, a continuous alphabetical list is far preferable, since in a subdivided bibliography the alphabetizing starts over with each section. Rarely should books be separated from articles, since a book and an article by the same author are best listed close together. It may be appropriate to subdivide a bibliography (1) when it includes manuscript sources, archival collections, or other materials that do not fit into a straight alphabetical list; (2) when readers need to see at a glance the distinction between different kinds of works--for example, in a study of one writer, between works by the writer and those about him or her; or (3) when the bibliography is intended primarily as a guide to further reading (as in this manual). When divisions are necessary, a headnote should appear at the beginning of the bibliography, and each section should be introduced by an explanatory subhead (see fig. 14.9). No source should be listed in more than one section. For alphabetizing, see 14.60--62.

I'm trying to find a style guide or similar on how to structure or organise a bibliography or reference section at the end of a book. Perhaps this applies most to a dissertation thesis where there are many different types of references. I am not looking for citation style guides on how to format specific types of references (I can find many such guides), but rather how to order them into groups by type, or even whether to do it in the first place.

For example, in my dissertation, I am currently considering a structure roughly like below (very simplified and inaccurate but only to get my point across). Now suppose I add more sections for journal articles, anthologies, etc... and this bibliography grows quite large, say over 30 pages or so. What is considered good practice here? See the two examples below. I find the first alternative a lot more tidy and easier to read, while the second is cluttered and just doesn't feel right.

I have found some recommendations which contradict each other. https://www.dur.ac.uk/resources/library/teaching/writingyourbibliography.pdf on page 3 says

You should not divide your bibliography into separate sections for different document types. References should contain all of the information required for a reader to find a source. Standards have been set for different document types to ensure that each reference contains the information necessary to aid retrieval of the source.

But there is no further explanation why there shouldn't be separate sections.

On the other hand, http://www.open.ac.uk/Arts/history/docs/thesis-guide-v51.pdf says on page 15:

The bibliography at the end should be divided into Primary and Secondary Sources.

And http://www.st-andrews.ac.uk/history/infoug/stylesheet.html says:

The bibliography should be divided into manuscript, printed primary, and secondary sources.

I'm writing as a historian/scholar and we tend to work with many different types of sources, so not just books and journals but also manuscripts, letters, promemorias..yes even postcards, maps, brochures, etc.

I'm open to hear what it's like in other fields as well and regardless of national traditions.

The three examples:

(1) Ordering by document type:

(2) Unpublished documents

(3) Archival sources

Letter from Ms. A to Mr. B

Letter from Mr. B to Ms. A

Other unpublished documents

Miss A. Diary notes.

Published sources

Public documents

Parliament records, 1980:CA-2:12231-13.

Parliament records, 1980:CA-2:71236-01.

Literature

Foucault, "The History of Sexuality" 1960.

Latour, Bruno, "Science in action" 1987.

Stevenson, Sue, "This is a book title", 2003.

2) Ordered by alphabet only:

Foucault, "The History of Sexuality" 1960.

Latour, Bruno, "Science in action" 1987.

Letter from Ms. A to Mr. B

Letter from Mr. B to Ms. A

Miss A. Diary notes.

Parliament records, 1980:CA-2:12231-13.

Parliament records, 1980:CA-2:71236-01.

Stevenson, Sue, "This is a book title", 2003.

Answer 1

The first person to ask is your advisor and/or committee members for whether dividing the bibliography into types is a good idea, and then if they've seen other examples. Next, check with the graduate school or other group that will be accepting the dissertation for publication. There is likely already a style guide and templates. Third, if there is still no clear guidance, look at previously published dissertations in your field.

Having a bibliography grouped by type could make it difficult to find references from the text. You don't really want to make your advisor or committee members do more work.

There's no reason that you can't keep a working copy of the bibliography sorted by type, just be aware that you'll likely have to convert it into one list at the end.

Personally, I can't recall seeing a bibliography sorted by type. I have seen ones divided by chapters or sections. I find this type of bibliography difficult to work with, as I have to find the chapter's list, then the individual reference.

Note: I have STEM degrees, but attended a liberal arts university as an undergraduate, so I have no graduate level experience in your field.