Review for NeurIPS paper: First-Order Methods for Large-Scale Market Equilibrium Computation

Weaknesses: After reading the authors response, I am convinced by the contribution of the authors. I really hope that they will include the answers that they provided to the final version of the paper because I find that it totally changes the reader's understanding of the paper. In particular, regarding the fact that obtaining a linear convergence rate is not obvious and was not done before. First, I found the novelty of the work a bit limited because not enough novelty was given in the "theory" side nor in the "numerical" side. The paper seems to "just" apply convex optimization methods to finding market equilibria and giving theoretical justifications for why this approach is sound.