First: I love ST. It is an amazing software that runs just as good on Mac/Linux/Win. In a pinch I will download and use it and never complain. Python extensibility is also great, since Python is the language I do best.
That said, and I hate to be this guy (not really), but... Emacs?
* Performance: Acceptable on modern day
* Goto Anything: Helm/Projectile, command palette = M-x
* Text Editing: It is what emacs does best as a program
* Themes: Yes
* Project Handling: Projectile
* IDE Features: These exist
* Source Control Integration: Many people use emacs purely because of Magit, it is a great git UI that people won't know about in general because it is runs on emacs
I copy my ~/.emacs.d to a server, literally my whole environment minus some tty nuisances just works. Everything. All my package versions. Emacs is stable enough that not many things break between big versions 23/24/25.
There are two areas ST wins: Out of the box configuration and performance on very big things. It is true the learning curve to settle all of these features I am mentioning is non-trivial, but Emacs is open source and, like I mentioned, very stable. It has gotten enough right in its initial concept that it will always be a fine environment to mangle up some text in files into whatever you want.
This brings me to my next point, there is nothing really new under the sun regarding text editing. Atom/VS Code/Eclipse/IntelliJ/Emacs -- you want to edit a bunch of files in a directory structure, navigate quickly and easily between files and get context important information about the code. Everything else is just flavors of those same chores, often specialized to a few programming languages or whatever new things are popular at the time (JavaScript / Can we build an editor in JavaScript).
I have seen very little innovation in the field of text editing. Light table was interesting, and its no surprise it is very Clojure/Lisp oriented. I am in a giant lisp machine when editing code in emacs, I can whack out a lisp expression anywhere and emacs will faithfully evaluate it for me.
Next point: If you work in a language or big project you need a way to navigate and generate lots of boiler plate in some languages (Java). IDEs help there, but...
Final point: The speed with which we write text as programmers is not the limiting factor of how much software we can produce and never has been.
My conclusion was to invest in one editor that has been around a long time and just not worry too much about it. I can use emacs in 20 years probably. Who knows about atom or vs code or whatever. So it isn't a pro emacs, you should use emacs reply to you, but that over the long run, using the same system to edit text is probably worth it. Stick with one thing and ignore everything else unless you are forced to use some IDE, and even then really try to avoid it. Even then I am pretty sure I could do my job with Gedit or Notepad without much of a productivity hit.
edit: formatting