Hi Raph, what's up? When it comes to improving yourself and learning new things, I think attitude is more important than tools. There is only so much you can get out of organizing your goals, resources and time if you overstretch yourself. I have found two techniques useful in terms of focusing my time and energy: Extrapolating from future achievements and elimination.
What I mean with "extrapolating from future achievements" is setting concrete goals in terms of where I want to be in five years, or what I would like to be able to say I achieved, and working backwards from there. I feel that the main reason many people engage with new ideas, technologies, tools etc. is the infamous FOMO, fear of missing out. We fear that we will be left out, worth less if we don't read this article or learn that programming language. If there is no actual driving force behind an approach to a topic, learning it will cost you a lot of energy. You will need to remind yourself again and again why you are putting in the time and effort, and even worse, the next shiny thing will be extremely distracting. If you start with the knowledge that it's taking you somewhere, however, you will have much more internal drive.
Elimination is just not doing things. You have three languages you want to learn? Drop two. Two books on algorithms? Drop one, or maybe even drop both and do some sports instead. I know this sounds silly; you are asking how you can get better at learning things, and I'm telling you not to learn them in the first place. But I think this is a key talent; dropping things and not looking back, not feeling bad about it, not losing any sleep over a missed opportunity. Everyone knows deep down that there is enough time only to concentrate on a couple of topics and areas in one lifetime; you can be a novice at many topics, but being an expert requires huge amounts of time and dedication. And the only way to bring these is by eliminating other topics. The previous technique of extrapolating from the future is useful here. Do you want to be called a great roboticist in 5 years? Then you will have to drop the ML. You want to be a great Rust programmer? You will have to let Clojure go.
I hope this is useful. What you have to keep in mind is that deep, multi-faceted expertise in a single area is very valuable, both as a trade and for you individually, to feel great about what you do. Acquiring this expertise is very difficult. You will need to put in a lot of honest work, will have a lot of dead ends and frustrations, and frequent doubts regarding your choice. Nevertheless, you should try to pick one area of expertise and eliminate all other efforts that don't contribute to your prowess at it.