Earlier quoted context omitted.
If an invention cannot be "reduced to practice", it is not patentable. My understanding is that reduction to practice means the invention has been actually built by the inventor, or that it is fully disclosed in a patent application such that a reasonably skilled person could construct the invention. If an invention hinges on some entirely theoretical concept that can not be demonstrated by even the most highly skill…
> ... some entirely theoretical concept that can not be demonstrated by even the most highly skilled people... In that type of case, I would expect the patent to fail the utility requirement (§101). But attacking it on written description or enablement (§112(a)) might work as well.
§101: https://www.uspto.gov/web/offices/pac/mpep/s2104.html
§112(a): https://www.uspto.gov/web/offices/pac/mpep/s2161.html