Writing the discussion section is always the scariest part of a paper or a thesis. I came up with some ideas to make it less difficult, but would like to hear other tips! Here are my two bits:
When writing an article, it is easier to start with the factual sections (methods, results) followed by the introduction and the discussion/conclusion at the end. It is important to identify the biggest gaps in the introduction, as you will come back to them in the discussion.
In the discussion, always start with the limitations: it is quite factual.
Then, focus on 3-4 main results that benefit from some additional context to understand what they mean, e.g. data from other studies or other countries, which can suggest some possible explanation.
Finally, try to relate these findings with previous theory or practices, and explain what your work adds that is new.
I agree with all your ideas, Maria; in my perspective, however, I would start my article with an introduction and pen down my methods afterwards. I think the IMRAD model is quite efficient. As to conclusions, I would just add that what pays out for me is writing conclusions in points (1, 2, 3, ...). In my viewpoint, this will make your results more accessible for the readers who focus on conclusion sections only.
Lead with your amazing results! Then discuss them in the context of other review papers, but so that readers will not have to turn to them to clarify what you mean. You do not want them leaving and giving attention to others lol. They may never come back. And never announce a finding, only to shoot it down :)