I planned to develop a text-base (?) game at the beginning of this year. In my first draft, it looks like this:

Then the Godot 4.0 alpha 1 released, I played it for several weeks. It was so buggy, so slow and crashed constantly. But the new sky shader feature was really really cool. So I revised my draft from the image above to this:

In past several months, I developed the game on both Godot 3.4 and Godot 4.0. Of course, the 4.0 one looks better for me. The shy shader is not only for background drawing, it also effects the cubemap. That means I don't need to design each mesh instance for different game scene. Just write one sky shader to each scene is enough. That is very helpful.
Now it is almost July. Godot 4 is still in alpha. Although it is much more stable than the first release, it still has lot of bugs and crashed each one or two days (at least in my personal experience). If sky shader feature backports to Godot 3.x, I will switch back to 3.x immediately. But it seems not in that case.
Anyway, since the project complexity is increasing day by day, I need to make a decision about should I continually stick in Godot 4.
If Godot 4 can become beta version before Christmas, it will catch my schedule. Otherwise, will not...
I guess no one can predict the Godot 4.0 beta release date, but how about you think? Do you the end of this year is a reasonable guess? 🙁