Our modules have cell-top BMSes (well, per module instead of per cell, but *very* similar to Yuasa) that communicate via galvanically isolated CAN. We also have contactors inside the battery box, so everything is completely galvanically isolated from each other, both on the comms and power side.
By the way, this is not something I should specify, this is absolutely necessary for any such system. The approach in the pics above is... well, I mean, I understand it from a prototyping perspective somewhat, this is done in the e-bike community, but it's not something you should do for more than a few test drives. 97 opportunities for something to go terribly wrong.
But to be fair, I don't know if that is what caused the issues. Battery fires generally happen only when overcharging or *generously* exceeding max cell temperatures. And even then, the (old Mn-based) Leaf cells endure 4.4V/cell and 75C apparently indefinitely. They are really, really good in that respect. Shorting them over those tiny wires is unlikely to cause any issues, even if you put, say, 100V excess on a cell, you'd just use those wires as a fuse.