Recently, scientists have began researching superatoms and supermolecules in depth. The premise is that existing, natural elements can be rearranged into clusters, in the lab, to exhibit properties they normal wouldn't - for example, "a siliconlike superconductor with the biodegradability of wood". Not only would these properties be unfeasible otherwise - but this substance has desirable engineering properties.
In a recent issue of Scientific American, this idea is listed as one of the "Ten Ideas That Will Change the World 2016" and is taken very seriously; this is not skepticism, this is a developing field.
What are the physical limits of this type of engineering?
Just how unnatural can your substance and its properties be if you explain it with this method?
Or, conversely, what are examples of the most extreme circumstances this can produce?
While I don't require a hard-science level of citation and calculation I would like actual scientific evidence. That evidence (or the lack thereof) can be cited, and thus this question is not limited to opinions.