I don’t want to dox myself, but I know what happened here. Performance was unacceptable on Series S, and it was unknown how effective optimization would be. BGS management is extremely risk averse, so they decided to add more loading areas. For example, making Neon multiple cells. Later the optimization team came through and it turns out that wasn’t necessary but CK workflow is way too manual and bug prone to undo the changes that close to release. I don’t remember if Purkey left before the buyout but Series S became the performance target and that made everyone freak out.