> Godot is much less featured but extremely easy to learn imo.
Also, for 2D, it is sooooo much easier to implement. No messing around with orthographical cameras, no tilemaps appearing with random lines, colours, etc. And 2D performance exceeds Unity's.
Does Unity have a true 2-D vector graphic engine or system of some kind? That seems like the main thing that doesn't look obvious to me as a total non-expert looking at Godot. (Mostly curious. I don't make games, but I hate the pixel-art 2-D style.)
Unity has an svg package which has been stuck in experimental state for years like many other unity features. I wouldn't call it true vector graphic engine. It can be configured to either generate bitmap sprite at build time (minor convenience over converting svg to bitmap manually) or it can generate triangle mesh (from what I understand also done at build time). In the second case once you start zooming it doesn't necessarily look very great either. Once you zoom out lines look bad since they are formed by thin strip of triangles which doesn't work too good for few pixel thin lines . And when you zoom in you start to see that its made from triangles instead of round curves since the generated triangle mesh is fixed. With both approaches svg benefits are largely lost compared to true vector graphic rendering engines.
I haven't seen any true 2d vector graphic engine suitable for games after flash beside the software made to be alternative flash players. Yes there are a bunch of games that still use that art style but in many cases they might be drawn as vector graphics, but then converted to bitmap for the game engine, maybe with some runtime shaders on top for sharper look.
SVG is just way too different from how game graphic APIs works. Just making a non real time SVG renderer is hard enough already. Even programs like Inkscape whose only task is drawing and manipulating are sometimes struggling with it. Open a more complex SVG and zoom in/out move around and you will see as Inkscape tries to keep up updating and caching the rendered image in chunks.
It might be possible to do something using browser svg support although I am not how good the performance would be with the kind of dynamic content games need.
There are also a couple of UI frameworks which support SVG. For something like that it can be acceptable to use slower but higher quality svg renderer and then caching result. For the most part UI elements are either static or at most move around due to scroll views and zoom rarely changing so caching works very well.
Also, for 2D, it is sooooo much easier to implement. No messing around with orthographical cameras, no tilemaps appearing with random lines, colours, etc. And 2D performance exceeds Unity's.