I devoloped my way of writing papers primarily due to my limited short term memory preventing me to being able to use the techniques I was taught, and so they focus on splitting the task up into small context-free chunks which produce a written result with me not needing to remember anything of it.
Depending on what exactly needs to be done the details can change but at a high level always consist of three main parts: “gather” where you collect a bunch of relevant information/facts (where each fact tends to be one line long, but sometimes more), “sort” where you take that mountain of independent facts and categorie, group, and sort them into a hiargy of information and topics (in which any specific information is easily findable and related information is clumped together), and “serialization” where you transform the hierarchy and facts into easy nice to read characters and sentences respectively (plus extra fluff like a inleiding which among others introduces the main concepts/structure of the paper (aka. the hierarchy plus a couple of facts))
For gathering information from papers (of which I already concluded they might be useful / which I have to use) I do so by first going through the paper(s) and copying (small) peases of text which contain useful facts/information and then afterwards making one sentence facts out of them with only the information I care about in them; these two steps could be done at once but they are every different tasks which thus is significantly mentally exhausting to be constantly switching between, plus this way you also got the quotes for the facts and with it a way to find exactly where you got your fact from.