I had a student come to me with essentially the same problem over two years and each time I helped her she was in refusal to listen as she stressed herself to just make it work now. Her problem was that she never took the time to do the basics and rejected any learning opportunity as it stared her in the face.
You get results over time if you dedicate yourself to just doing the thing. For many subjects there is no shortcut, no way to walk the path without actually walking it. Every time you encounter an issue there is a learning opportunity. Use it.
You follow a tutorial to do something, feel happy about it. Then you start a new project to put your new skills to good use and... Blank. No idea where to start, no idea how to proceed.
It's so important to build stuff, using references is fine, but following tutorials is not the way forward! You have to work on your own without the training wheels.
For me it is. Even in my domain where I’m an expert and it’s fun, it only is if I’m working on something interesting.
Sometimes distraction is the main issue when it comes to having ideas.
I get stuck into this mentality of "I need to learn and master X, Y and Z before I can even start building my dream"
Would be much better served by just building whatever and learning the skill
Once you're a little more confident (you know a bit, but not much) I suggest to modify the tutorial as you follow along, that makes the tutorial harder and gives you small challenges to overcome while still giving you general guard rails.
Then as soon as you're dangerous enough to be let loose you should pick your own projects that are slightly above your skill level. Maybe try different approaches if you're unhappy with the first result.
When I wanted to improve my comic drawing skills ca. 2009 I started drawing and publishing a daily webcomic strip for a year. That really helped.
But tutorials remain useful even if you're advanced or a pro. E.g. if you use blender a lot and a new feature comes around watching a tutorial on it is a very efficient way of getting up to speed. Of course you will watch tutorials differently from a beginner, you will pick up on different things etc.
The best way to learn is a serious project with a deadline, but if you have that deadline it will make you wish you had watched some tutorials when you had the time. Source: I teach this kind of stuff at the university level for 6 years now.
In my experience most people can do this, if they think about it a bit — identify the thing they want to learn and find a tutorial for it. Which is amazing, really; this sort of meta-knowledge is a remarkable human concept.