A canonical example of this is gallium, which famously doesn’t concentrate or form ores. However, the process of refining aluminum coincidentally partially refines gallium as a byproduct. So almost all gallium is produced by continued processing of the aluminum refinery waste stream even though aluminum ore contains no more gallium than a random rock.
China produces almost all of their REE from secondary and tertiary ores. The prerequisite to having these secondary and tertiary ore process is having a primary ore. If you are not processing primary ores, none of the secondary and tertiary ores will be available to you as an option. If you want to have a supply chain for diverse metals, you need to be processing diverse primary ores with an eye toward reprocessing the waste stream when it is chemically efficient.
The US has outsourced much of the primary ore processing that can produce a lot of metals that can only be economically produced as secondary ore products.
Of course the ponds create lots of waste gas and water, including radioactive elements and metals since you end up dissolving and concentrating those also. (So both metallic and chemical contamination) The ratio is something like 1 ton of rare earths = 2000 tons of toxic waste.
Seems to me like the issue is not so much the capex cost, but the regulatory and environmental cost. Doing it at scale in a way that doesn't harm the environment (and proving that) is likely prohibitably expensive in the US.