GNU Emacs is basically the Right Thing, and we just need to keep improving it and adding new stuff. A fork would be a disaster, because we are trying to unify the Emacsverse.
There is no point in trying to support any other flavor of Emacs than GNU Emacs 22-and-up, especially if we want to take advantage of all the new stuff (multitty, network servers, etc.) I want a single target platform, not the Java-esque “if I’m in xemacs, do such-and-such…”
It has a three-decades-deep archive of working code, and a thriving development community today, so it isn’t going anywhere. My solution is 1. develop workarounds for, and raise awareness of, what exactly these awful problems supposedly are with Emacs Lisp, and 2. promote cl.el and object-orientation.
We all know about this one.
The problems we encounter will point the way to (and show the need for) changes in Emacs itself.
I want ELPA to integrate with the Debian packaging system somehow
We need to improve the display engine. There was something in the Emacs TODO about emacs being able to absorb an X window within one of its windows.