If you haven't found them yet, I'd recommend watching Grant Abbitt's videos on YouTube. There's years of tutorials there, plenty of them regarding low-poly models, so scroll to the bottom and then Ctrl-F for what you're interested in. For example, here's the first of a four-part series on modelling, texturing, etc., a tank.

Ultimately, with anything technical you are going to have to learn technique and, with practice, you'll stop feeling out of your depth. The just-in-time vs. just-in-case learning dilemma will only feel worse if you try to run too fast before you can even crawl—throwing yourself at the steeper part of the learning curve can hurt—so, before jumping into the thing you really care about, make some simpler things that you're less invested in for their own sake.* Instead, as you practise, think about how you can apply it to the thing you do care about, and have that be the motivation for working on them. It doesn't take that long to pick up the basics and, once you do, you'll be in a better place to accelerate your learning in the specific direction you really need, because you'll be in a better position to recognise what will and won't be immediately useful to you. In the meantime, you don't have to put off the motivating project. Work on it. Refine it with each new thing you get a handle on. Soon enough you'll have something you want to show us all.

I end up having to manually do a lot of fine-tuning to make the process work with what I'm doing.

Yes, this is what makes it your work. Pretty much all creative endeavour involves combining and modifying simpler building blocks.

It's not "all the hard way in the end", but there is no end point other than the one you choose. There'll always be something new to learn, some improvement you could make, so start simple and build from there.

* This makes problems and mistakes less demotivating because it's easier to drop and try something else.

This motherfucker just learned about select sharp edges and mark seam ahahahahah

one more thing down.