My understanding of the situation is that Ernest, the main developer behind Kbin, thinks of the current Kbin as a proof of concept, and he is doing profound rewriting of the codebase to better fit his vision of how it should be working.
Meanwhile, other people wanted to contribute to Kevin directly, developing a better product on top of what Ernest considers to be too shaky foundations. So he's not all that interested in pursuing that part of the development before he is happy with the core.
This also leads to a dynamic where he still has his own vision for the project and it goes through him, whereas other contributors want to make it their own more and develop something different.
It's hard to see how to make everyone happy here without forking. Hopefully both projects can still gain from each other in the future: Mbin can benefit from the rewritten codebase of Kbin, and Kbin can implement features from Mbin after seeing that they are good and work well. In either case, the continued development as separate projects is probably not all that bad.
testing