The only thing that gives bitcoin value is the proof of work. It is what solves the distributed, trust-free consensus problem. It is a solution that should not exist to a problem that just a few years ago smart people assumed to be unsolvable. Asking "why is mining useless?" betrays a certain ignorance in the utility of proof of work. I suggest doing some more research into it.

In particular a proof-of-work function needs to be self-validating, it needs to be universally available, it needs to be progress-free, it needs to represent economic cost, and surprisingly it needs to be intrinsically worthless. If these conditions are violated, it fails to be useful as a consensus mechanism.

Hash-based proof-of-work is extremely useful. It gives economic assurance about the cost of disrupting consensus.