No, neovim is still the same concept: you're scripting the editor. It's still this gargantuan thing with all of these things you bolt on and reams of documentation to wade through. It still has the vim philosophy (which is the same as the Emacs philosophy, only not as good): my editor is an operating system and I want to rewrite all of my software to run inside of it.

I'm thinking of something altogether different: you leave the editor alone and you script the operating system around it, using the editor only as a flexible control interface for everything. This is the original vi philosophy. It means you don't have to rewrite anything. Any new program you install, if it can send/receive text on STDIN/STDOUT/STDERR, then you can use it in vi.

So what's wrong with vi? It has a few shortcomings: no unlimited undo, no ability to pass the cursor (row,col) position to an external command, and working with multiple buffers is more painful than it needs to be.