IMHO, the next step for BBR is that CZ changes (simplifies) the code base such that it could be (easily) taken over by some group of interested devs. Right now, the source code reads unready compared to, e.g., XMR code. There were also a couple of problems, when I remember correctly, with compiling the BBR code. The latter is probably the most important thing that should be improved. BBR code must compile for everyone on Linux and Windows and, maybe, MacOS too. When that stage is reached (maybe it is already reached; but not to my knowledge), BBR can start to evolve. To me, BBR looks like a frozen fish right now that either needs to be eaten asap, or revived and put back into water. Of course, I am in favor of life