Love2d is so satisfying... I worked in game maker for years, and I think I was unnecessarily intimidated by "real" game programming. If you're used to something like game maker, a lighter weight framework can be basically the same thing, just with some extra steps. Instead of letting game maker handle collision, I write a little collision function/loop (for my games it's like ten lines of code). Instead of using game maker's built-in level editor, I use Ogmo to generate a JSON file (containing the tile and entity placements), then use a simple library I downloaded to turn that file into a Lua table, then write a function to draw everything on the screen.
It's a little scary at first, but it's just ultimately not that much harder. All the hard creative stuff, level design, storytelling, sound direction, etc., that's all stuff I already had to do in game maker. Working in Love2d is the same, it just takes some extra time to get the initial prototype up and running.
And none of this is unique to Love2d. If I want to try out Haxe/Heaps, or Raylib, or any other number of frameworks/lightweight game dev environments, nearly everything I've learned in Love is transferable. I don't feel beholden to any one environment anymore, which is really freeing.
Game Maker is only going to get more restrictive and shitty with its licensing. It's scary to think that someday I won't be able to get my GMS1.4 license running on a new machine and I'll be completely locked out of my old project files (unless I want to mess around with cracked versions of course, which aren't going to get easier to find as GMS1.4 gets older and older). Working with FOSS like Love2d and Ogmo (or any of the Haxe frameworks, Raylib, SDL, etc.) helps me feel safe from getting fucked over by uncaring corps like Yoyogames and Unity and Epic. I think gamedevs are becoming increasingly aware of the value of that lately.
Necessary addendum: specific, high level tools like rpgmaker, ren'py, puzzlescript, bitsy, twine, inform, etc. are wonderful, and I love them to bits. (I do not consider game maker, unity, godot, unreal, etc. to be high-level tools, they still require you bring a LOT to the table yourself.) Genuinely high-level tools are simple and straightforward enough that you can sidestep most programming headaches and get straight into the creative work, the storytelling, the level design, the direction. The "negative" is that compared with from-scratch gamedev, the tools define a lot of your game's voice and game language for you. And that's totally fine, because there's still a ton of creative work you can do within the natural confines of what the engine is capable of.
For many people, the limitations of these engines result in them making more and better stories then if they tried making stuff from scratch. That's certainly been true for me. I love Love2d, but I'm always going to cut my hardcore programmer-y gamedev with work made in high-level tools, tools that let me jump straight into the creative work without mucking around in the game logic for ages. That was true back when I made games in game maker as well.
But if you're already using game maker or unity, then you're already doing a ton of "programming headache" stuff. You're having to mess with the tech, the game logic, you're not getting straight into the storytelling. You're probably doing that 'cause you enjoy the programming logic stuff (I certainly do), and/or because you have a very specific vision for your story. And if you're already there, I think it's a much shorter hop and a skip over to stuff like Love2d or other frameworks than people realize.
(For 2d stuff at least, I don't have the faintest clue how 3d gamedev compares between different toolsets. This is fine though, because 2d games are better than 3d games and everyone should make 2d games exclusively.)