Good writing is about crossing different sources to get closer to reality, so note-taking—the act of recording information—is the writer's best friend.
Note-taking is an integral part of drafting. In this sense, the most prolific writers are often the most prolific note-takers. Most of Kafka's work was unfinished. Proust's notebooks were incredibly chaotic.
It is indisputable that becoming a better writer implies taking notes. The million-dollar question, however, is how one should go about it.
I already wrote about my note-taking system, a mix between the Cornell note system and the SQ5R methodology.
I just use a plain text file to keep my notes together. I know that note-taking apps are all the rage right now, but I think using one is missing the point: notes should remain transient, by definition. I either edit them as soon as possible into publishable content or get rid of them, just like I would manage an email inbox. I refuse to store and structure them in the long-term: notes are not the end-goal, they are a gateway to bigger bodies of work. Notes taking the dust are worthless and clutter the mind, in my opinion.
I also use voice typing. I recently invested in a Lavalier microphone to take notes whenever I can. I connect the mic to my smartphone and use a dictation app. This way, I can just sit on the couch, have a walk, or sit on a park bench outside and let my train of thoughts take me away.
Switching to a voice-based workflow really helps me improve my productivity as a writer. Once I'm done with my voice typing session, I export my text file from my phone to my laptop, go through it, and divide the notes into two categories. When there is enough content and I already have an outline in mind, I just create a new draft, insert the note, and add a bit of structure. On the other hand, one-liners are inserted at the top of my plain text file centralizing all my notes on my laptop.
Every weekend, I go through this text file and do my best to reach the inbox zero state I mentioned earlier. The best notes/ideas eventually come back. Most ideas are garbage (that's ok). All of my associative ideas come from actual published content and conversations, so I don't fear missing out on those either.