Not so much a lesson learned as a lesson reinforced: there are still too many gaps and inconsistencies in data sources like Wikidata and DBpedia to rely on them for robust discovery features. Examples of this include the publication timeline and influence diagram that we included for authors. The inconsistent use of IDs across these sites is also a drawback, forcing the use of multiple queries or queries with unions to try to find matches or making locally stored IDs/URIs less relevant.