Also releasing a tool that you hope other people will use means you have to actually support it: update it, bugfixes, etc.

and to pass the "dogfood" test, you'd need to use them to build games as well.

Recently I tried to develop a level editor for my small homemade game engine. As you know stuff like this is very time consuming and in most cases the result compared to the commercial soft wares is poor. After hours of work, using different developing packages, I came to an Idea and that is about developing a general purpose editor.

An editor for your engine should not be "general purpose"; if it is, it's a distraction. Your task (making levels for your engine) is limited, and rather than a powerful single tool you should develop a workflow that consists, as much as possible, of reused existing software: for example if you have simple 3D scenes you can write an exporter plugin for a general purpose 3D editor or a converter from one common 3D model format, and if you have scripts you can write a command-line compiler and syntax highlighting rules for an advanced text editor.