Quantum clearing a plain Turing machine on cost is the smaller claim. A practical NP-solver has to buy an exponential search with polynomial resources, and thermodynamics usually sends those proposals back. If one physical resource in the review still stands after noise and energy, that is the argument worth having.