Right, first of all, while I’ve read papers for FFG and CBC, but it’s been a while now, and I may not fully understand both protocols and all of their pros and cons, so perhaps after I finish gossipsub I’ll scrutinize them again. I like that it’s correct-by-construction, i.e. using maths to formally prove safety; that you can explore the full tradeoff triangle (which may make further R&D to see if you can have different applications using variants of the protocol designed for a different point of the triangle interesting), extending on relative local views of safety from each node to consensus safety, etc. But perhaps I shouldn’t be so adamant about which one to focus on implementing in the short-term, and may need to look more into it to get a better idea.
But generally, like I’ve said, it seems that it would be less work in the long-run to go straight to implementing CBC.