ATX dying would probably require a drastic upheaval beyond even CAMM.
Likeliest candidate for the executioner would be tied to development of glass PCBs with photonic interconnects – something I don’t expect to land anytime soon even with TSMC and other non Intel parties making moves on photonics.