Ah… Ancient Computer User here. When Word first introduced the concept of Styles way back in the 70’s? 80’s? it was the most confusing thing I ever ran into. Scrivener is much more intuitive to me; nothing has a “style” unless I tell it to have a “style”.
The reason for this is Scrivener’s intended usage. Scrivener did not shine when I tried to use it as a “What you see is what you get” word processor. I tried this when I first switched to Scrivener from Word, and I was almost as miserable as you are now . I had a devil of a time when it came time to compile a document for both a) Web display, and b) a fully-formatted PDF for an accompanying handout.
By design, formatting in Scrivener takes place as much as possible in the compiler. That’s how I have ONE primary text, and can format it easily for generic ePub for upload to services like Kobo, Nook, and iTunes, a specialised Kindle format, a simple old-fashioned Word .doc for Smashwords, and a fully-formatted book interior PDF for CreateSpace. Formatting happens after writing, not before.
I suggest you don’t put titles inside your documents, nor sub-titles, nor sub-sub-titles. Instead, I suggest you break your documents into smaller pieces, nest them in the Binder, and name each document with the (sub)(sub) title you currently put into the text. When you compile, you may then format those titles as you choose, hierarchically if that makes sense for your overall project. Only use styles for either a) formatting within paragraphs (italics, bold, etc,) or b) special paragraphs such as block quotes. When it comes time to compile, you may (and probably will!) override ALL such things as fonts, paragraph spacing, indenting, etc. for each style, the document titles and for the “no style” body text. Think of “styles” not as styles as Word uses them, but just a way of denoting to the compiler, later, that it needs to handle this chunk of text differently.
Currently I use a monospaced font in the editor for my default text that by design makes every typo as painfully clear as typography can make it. In my compiler formats, that gets changed to various other fonts (Courier Prime for distribution to beta readers as ancient as I, Times New Roman for submission to less antiquated editors,… Baskerville for my CreateSpace PDFs). I also use line and paragraph spacing in the editor that are quite different from my output formats, again because, in the editor, I want to make it as easy as possible for me to read on-screen and to find my paragraph breaks.
In short, I suggest you view formatting in the editor as that formatting which is comfortable for you WHEN WRITING, and also contains enough information for the compiler, later, to format your document for distribution. It is VERY much a two-step process by design, and trying to do it all in the editor only gave me grief and frustration.
Hope this helps.