I have really two answers to this question.
Since my principle storage is electronic. I'll start there.
How you organize your electronic folders may relate to how and how much you actually save. I try to save things that are not readily available elsewhere. This means that I store few census records, and now fewer and fewer marriage and death records. I do store pension files, deeds, deed indexes and probate records. Unless someone has sent me the paper materials, I don't set out to store any of my research materials in paper form. I store all of my research reports in electronic form. I used to send them broadly to cousins (ala, spam the family), but I find myself blogging much of that work now. I save many e-mails to pdf.
The computer filing system I have used since the 1990s is summarized below.
Note: I have patronymics in my family tree and the spelling of some names changed over time. The word "surname" doesn't quite describe top level organization; here I'll use the term "family lines." Hopefully the examples below are helpful.
Top level organization: Each of my ancestral family lines (as in a pedigree) has a high level directory; I call each a "log." I prepend these folders with the word "log" (including an underscore) so that these folders will group together and can be forced to sort at the very top of my Genealogy directory. For example
Each of the family lines has a default subfolder system that classifies a "generation," and each of direct ancestors has a first level folder. I assign a subfolder to the other children of my ancestors (siblings). For this purpose, I use a numbering system that is not unlike Register or NGSQ to track each "node" of my direct ancestors in the family line. An example follows.
Other than for the higher level direct ancestors, I don't create a folder for the other children (siblings) unless I've actually conducted research about them.
For every surname, I have a "dump" folder, too. I use this for information about the larger family and for things I don't want to categorize further at the time. These "dump" folders carry the same general name, "XXX [surname assigned to the family line] and Related"
From time to time research turns into a what for this purpose, I will call "negative proof." From time to time, I create "Not" folders to hold that research. So, for example:
Under the main folder/directory, "Genealogy" I have a number of other folders. A few examples follow:
- Genealogy General. This folder contains electronic version of my genealogical reference materials. For example, in this folder and its associated subfolders I have Black'sLaw, Evidence Explained, Numbering Your Genealogy ...., etc.
- Descendant Researchers. This folder contains family files shared with me and some e-mail histories. Usually there is a subdirectory that carries the name of the collaborating researcher. For example, probably the largest folder in that section is titles, "William Smith," who I've been collaborating with since the 1990s.
- Album. This folder includes the work I'm doing to digitize family collections.
Somewhat on topic, I blog and work to develop full citations for the entries in my family file. I take full advantage of the note keeping features of genealogical software. The benefit is that today I don't really obsess much over the drive organization the way I did in say 2000. If I have recorded the author and title of a source in my file or on my blog, I can almost always find that puppy and all my related related research notes.
As above, I don't set out to "store" any materials that are readily available from "stable" sites on the internet. I may download a birth or census file for the purpose of attaching it to an e-mail, or use it to create graphic for a blog article. I'll more than likely delete that electronic file after I've finished that work.
I do invest some worry time in the material paper files and collections that I hold, as these generally represent privately held materials that have not been widely circulated. I work to keep those materials in tact (even in the original order received); most are stored in archival format (archival sleeves, etc.). Dictated largely by dominance, those materials are generally organized by surname (family line), researcher name, or collection/repository.
When I digitize collection materials, each item (scan) is given a descriptive number or code. I write the code/number on a sticky dot that goes on the face of the archival sleeve. An example of a code is "600T-1063." The first set of code represents the resolution and file format (so 600T stands for 600 dpi, Tiff). The second group of numbers represents the electronic scan number (part of the file name; ultimately intended to be part of the metadata).
Update 2: You asked specifically about where to store birth information about children--whether with parents information or the folder about a married adult. I don't think I really have a black and white rule, but my direct ancestors families are organized in the male ancestor's line--so most of the records about my maternal ancestors are stored under their husbands' family line. Research about my maternal ancestors sister-siblings, however, are stored under the father's family lines.
There are well documented paper storage methods that recommend storing information by record type (birth, marriage, death, deed, probate, etc.). That system didn't work as well for me, maybe because I have patronymics and, even when I don't, there are so many similarly named persons in my different family lines.