Like many people, I’ve spent the last few days looking for the new Unity. Godot has some potential, especially if it can take advantage of an influx of dev talent to drive rapid improvement. Open source is cool like that. However, one major issue holds it back - the binding layer between engine code and gameplay code is structurally built to be slow in ways which are very hard to fix without tearing everything down and rebuilding the entire API from scratch.
TLDR - Godot’s raycasting is considerably more memory intensive than Unity’s. Using the Godot API to call its raycast in any language is much slower than simply using the builtin Node. Most of the performance woes are due to how GDScript is implemented: a slow, interpreted language (Python)
It’s a fair complaint, Godot is far from perfect in many areas and may not be ideal for certain projects. This is fine, really. If it’s possible to make the engine better, I hope the pull requests that do so get accepted
He did a small complaint that GDScript should be ditched. If you care entirely about performance, that is true. The problem, obviously, is that a lot of people, me included, rely on it instead of C#. For all intents and purposes, this is something that could be done, but it’d end up as a wholly separate branch of Godot that would end up with several incompatibilities, eventually becoming a different engine. In that case, it might be better to just check an alternative like Stride3D or Torque3D