For me, the code doesn't make it a hard project to join. It's a hard project to project to join because lb-refactor is so much more than a refactor and the fact that master is so far behind. Because of this, it feels like a) it's not a community project, and b) features that I might want to add are going to conflict with whatever might happen next to lb-refactor. It just doesn't feel like a stable project to join.
1. I have an open pull request for pulling lb-refactor into master
2. You can base new features on lb-refactor, or if you don't I will resolve the merge conflicts myself
3. Are you saying that it's worse to have a small start at using travis-ci than to not even involve travis-ci at all?
I too have to admit the project is hard to get into. I'm having to spend a good amount of time studying the code, taking notes as I go. I feel if the project had more documentation in general, it would be easier to get into.