There is no known best way to improve. Besides, everybody is different. There are many ways up the mountain.

There is a lot of advice out there. Often people repeat what they have heard, or say what worked for them. OC, what worked for them may not work for somebody else. Everybody is different.
If there is a sine qua non, however, I think think it is this. Review your own games. It is good to review professional games, but even better to review your own games. First, they are the raw material you have to work with. They are what you want to improve.

So do the work. Second, you have already spent some time thinking about them, as you played the game. Your thinking is also something you want to improve.

Now, as an SDK, you have probably already learned some bad habits. That happens to everybody. You have to root them out. That is not so easy. Unlearning is in general hard to do. Besides, when you learned the bad habits, you didn't think they were bad. If anything, you thought they were good. And you still may think so.
One thing I have observed in SDKs is sticking with what they have learned. A little knowledge is a dangerous thing, as the saying goes. They will make certain plays, or certain patterns of play, whether they are appropriate to the situation or not. Sometimes people will avoid a play because they do not know it. They don't know how their opponent will reply, they don't know how to follow it up. They stick to the bright path and avoid the dark forest. If you stick with what you know, or with what you think you know, well, that's not the road to improvement.
In your reviews you have to question yourself. Take nothing for granted. Is there a better play somewhere else? I made small life. Could I have sacrificed those stones, or some of them? What if my opponent had played there? Be thorough. Be objective.
Fortunately, with AI we have very strong players to help us explore our own games. AI suggests a play we don't understand? Great! That shows a gap in our knowledge that we can start to fill in. Or at least explore.

It's not like we will fully understand it after a few moments of thought or exploration, playing a few variations on our own, but we will understand it better. We will have learned something. Which, after all, is what we set out to do.

Bonne chance!
