On Swamps and Morasses
There are a couple of schemes that come to mind immediately that might provide some grist for the mill, if not a full English breakfast.
First, of course, is the one we already use! You assume the existence of sentences and paragraphs already, but I think we should quickly look at them for efficiency sake. We generally learn to write a text, like an essay or a thesis, beginning with the thesis statement, the sort of summation of the whole problem the essay addresses. This is considered to be the statement a reader is to comprehend first. Next down from the entire essay in hierarchy is the paragraph. Likewise, we're taught to begin each one with the main sentence, that one that introduces the main idea to be discussed or revealed in that paragraph. Lastly in the hierarchy are sentences themselves. Interestingly, we don't order sentences "logically" but rather grammatically. Even though one might write a dissertation in Latin and put the thesis statement first, the sentence itself might not actually get to the point until you get to the verb way down at the end.
This scheme is both Linear in the spatio-temporal sense and Top Down in the hierarchical sense.
Secondly, we could consider extending this basic scheme by standardising and increasing the power of punctuation and section markings. Many technical documents do this by noting every paragraph with what amounts to a unique serial number. Often in historical grammars we find references to things like P4a.2.N3 which denotes paragraph four-A (a subset of lesser importance than para 4 itself), section 2, note 3. All in all , a very low importance reference as it's relegated to a foot or end note.
We have lots of interesting, unusual and unused punctuation marks that could be regularised, standardised and ammended to fulfil several key roles from marking importance to speculation or editorialisation. We have things like interrobangs and exclamation commas and the like; some languages have sectional punctuation. All of these could be incorporated into the existing scheme relatively easily.
Thirdly, we should take a look at other systems entirely different. First, let's take a look at music and in specific the conductor's score. Unlike an individual musician, the conductor has access to all the parts that are playing at any moment in the piece of music. Very much like what you're looking for, the conductor needs to be aware at any moment of "what is to be comprehended first" and "how does the whole text flow". Although music itself is linear, and the most important theme or idea isn't necessarily the first thing you hear!, the reader of the score needs to comprehend all the different sections, moving parts, thematic ideas and when they all occur.
As you look in the images in the link, the conductor has marked up the "text" of the music by coordinating various divisions by colour, notes when various themes, motifs and entrances occur and the like.
Even in an individual musician's score or in a piano work, you'll find plenty of "flow" markings: dynamic changes, pedal settings, phrasal markings, accentuation types as well as stylistic & tempo markings.
Some types of music rely heavily on flow marking, for example the genre of unmeasured music. Melodic and harmonic flow are indicated by the composer through the use of sweeping and flowing lines, but the interpretation and execution of the flow is the work of the player.
Another system to look at is hypertext. We've all used it when we hover over highlighted key words and clicked on links to supporting data. And we've all wondered why the composer chose to link two things together that don't make sense just because the other word is spelled the same.
This system allows the writer to make use of the linear nature of reading to ensure that more important topics are marked and read first, but also allows freedom to move between sections and also to dive deeper into a topic through the system of links and hovertext.
Wiktionary puts this to good use by ordering the Important Stuff on the main article page, while linking to supporting data, historical changes, usage quotations, and hiding grammatical apparatus and the like in spoilers.
Take at-baill for instance, an Old Irish verb meaning to throw. The key points and flow are handled here on this primary page with its sections delineated (word, etymology, pronunciation, parts of speech, declension/conjugation, descendants, etc. Clicking on the spoiler bar opens up a whole new view into the agony of grammar. Especially for anyone who thinks Greek verbs are truculent.
The links in all the sections lead one down rabbit holes of time and space and history into other languages, other peoples' perceptions, related forms, semantic drift and all kinds of other interesting inquiries that don't have anything at all to do with the basic thesis you started out with!