These only holds if you 1) think US public services should divert precious human capital to satisfy your agents 2) that this infra already exists or can be built swiftly 3) that somehow removing these frictions are a good thing at breakneck pace. The cited paper doesn't even say this.
A consequential choice will be made soon by government and companies as they are increasingly flooded with agentic AI inquiries that overcome the frictions that gated access (see the graphs - it has already started!) Two options: do you make services worse or better as a result?