I'll add onto that that Unity appears easy on the surface and the tools are really good, but it never really assumes you don't already know what you're doing. It expects a level of understanding about game development in general and that you can make your own decisions in how you want to make your game. It's a blank canvas. That's why I really like it, because I don't like having my hand held. But I don't know how easy it is to use going in with little to no experience. I'm sure it's fine but I can imagine there's a lot to take in when you're starting out.
Comparing Unity vs Ren'py I find is the same as talking about Blender vs Daz3D
One program is specialized, kind of like a power tool, it does one thing, but it does it really well.
And the other program is more like a tool box, there's a lot there, and a lot you
could do, but since they are more of a 'general' toolbox/playground it's not really
structured for any particular purpose, without structure it makes it hard to get started because you have to make your own structure.
Unity VS UE4 is different. both are full on game engines, but UE4 was made with FPS games in mind. While UE4 has an amazing graphics pipeline, the engine can kind of get in the way of what you want to do because it's structured for FPS games, so you have to work around it's structure to make different kinds of games. Now for some people this is easier, adapting from one foundation you know (such as people who use RPG Maker) and modifying it to fit your needs (like how people use RPG maker to make non rpg games), since it can be easer to modify something than having to start from scratch, especially if you've never built custom game systems before.
The only real reason to jump into something like unity or non specialized engines, is for when you have an idea you really want to bring to life, but after investigating, turns out there's no way to implement by trying to modify something. For example, those interactive phones or minigames in ren'py if I understand correctly, are examples of a complex game system that ren'py was not designed for, but that clever people managed to get working anyways. Meanwhile, if I want to have some complex AI operation in the background that is asynchronous to what the player is doing (eating lunch away from keyboard), that's not something most specialty engines can handle, I need to work with raw code to allow for parallel operations to occur while the game is running.
For me, unity wins. because for me, it uses C# which I think is the absolute best language to 'program' in, its the most documented game engine (the main issue with Godot is its still new and undocumented and unity pretty much copied most of their good features), and I really like its supplementary unity math library, and the canvas RectTransform for making very flexible user interfaces. If it wasn't for the RectTransform and WebGL export, to allow for me to quickly get started with more complex projects with the confidence to know it will be multiplatform ready, I'd actually use MonoGame Framework to code my projects with, but that's raw code, no nice user interface.