Maybe for e.g. RHEL it makes sense to maintain Python 2 beyond end of life, since they can charge an arm and a leg for companies who have been too lazy or complacent to port and upgrade their stuff, despite clear and generous deprecation timelines. But I don't buy this for Apple.