Paper sections, hunh, what are they good for?

Phew, it took a helluva lotta time, but the twin attention grabbers of finishing/submitting our manuscript and then the U.S. election have finally subsided. And that's good, because I was going crazy, first obsessing over the details of the paper, and then the daily election polls.

Working on the paper made me realize that I am simply crazy about some things. Like, I'll argue six ways from Sunday against abbreviations, which any sane person would just let go of. And that there are some aspects of a paper that I will work for hours upon end to reach some personal Platonic ideal, whereas there are others that I couldn't care less about. That made me think more about the goals and importance each section in a paper (e.g. Abstract, Intro, etc.). Coincidentally, today Female Science Professor posted about Introductions.

It made me wonder how other people rank the importance of paper sections. For me, I would put it as follows:

• Figures. It's the data stupid! Figures have to look good, and embody the points you want to make. I love it when a good figure whallops me with its significance. I'll admit that this is damn hard, and I can't be sure my own figures always meet this criterion. And, it's not simply the visual design of the figure; it's also a combination of experimental design, clarity of the data themselves, and the placement within the flow of the argument. This is perhaps obviously the part I spend the most time and effort on.
  
• Abstract. I put this here, even though typically the amount of time and effort spent writing the abstract isn't terribly large. But it's so important IMO for two reasons: first, it's indexed in PubMed, so beyond the keywords, this is a place to put terms you think your target audience will be searching. Second, it's a place to put the one or two take home messages that you want the reader to come away with. Something short and sweet that will stay identified with your paper, perhaps a nice number indicating the size of your effect.
  
• Results. Speaking as a reader, I may not always fully read this section, but if and when I do turn to it to evaluate one of your experimental findings, it'd better be super clear. Usually I want to know why. Why did you actually do this experiment? Sure, it might be obvious to the authors, but it's not always to the reader. So tell me. Throw me a frickin bone once in awhile. Then tell me what you think the results mean. Don't leave the interpretation to the methods.
  
• Materials and Methods. Again, these are required to really evaluate the experimental findings. And by the gods, if you use a method but mention it solely by referencing one of your old papers, then you're likely to lose me. It's ok to reference yourself there, but at least give me a brief sentence to describe what you're doing.
  
• Discussion. Try and tell me where your stuff fits in with the previous literature, and extend what you think it might mean. Suggest new experiments.
• Introduction. Turns out I differ from FSP here. I like a nice short and sweet introduction, with just the barest essentials. But I wonder if this is a disciplinary dependent effect: there are so many reviews published in biosciences now, that it's easy to reference those as a stand in for any more detailed consideration in a single given paper. Is that the case in physics?
  

Actually, this list make perfect sense to me. I have always considered myself very datacentric person, in that data has the highest priority over theories, models, frameworks, whatnot. And, this list matches the order of sections when I read papers, and the importance I give to each. For example, if I'm pressed for time, I might not read the Intro or Discussion at all, or maybe just read the first sentence of each paragraph.

Still though, maybe other people have different lists, and if so, perhaps my papers are failing them? So let's hear it readers!