Greg KH has given a few talks/blogs about the design goals of kdbus. https://lwn.net/Articles/551969/
Another one?
If the existing solutions were sufficient, why would the 'new' way ever get merged into mainline?