The aim of Wikidata is
- "Wikidata acts as central storage for the structured data of its Wikimedia sister projects including Wikipedia, Wikivoyage, Wikisource, and others."
I have done my best to have Wikidata hold useful data on taxa and on scientific names. A database can never be better than the quality of its data. Being a "collaborately edited database" Wikidata is very vulnerable to mass imports of stuff found somewhere on the internet. A second is the handicap that any error made in a Wikipedia will be imported into Wikidata.
When it comes to taxonomy, the focus of Wikipedias is on species, also genera, families, orders, etc., in short on taxa (groups of organisms). So for Wikidata to serve in support of Wikipedias, it has to hold data on well-known taxa, their possible names, etc. Within one Wikipedia, in general, a single correct scientific name will be used for a taxon, but different Wikipedias will differ in what this single correct name should be. The real world may hold even more variation on what this single correct name should be.
Years ago, Wikidata adopted a model where each possible name of a taxon gets its own item, this item to hold the nomenclatural details of that name and the data on the taxon involved, if published by that name (that is, information on the taxon by a different name will be stored in that other item). So each item is on both a particular name and on the taxon by that name: it should gather statements on both.
This left the question of what to do with scientific names of the kind that cannot possibly be taxa, never represented taxa, and never will represent taxa. Attempts to develop a datamodel for this were not succesful; this is not exceptional as about half the data that Wikidata needs to model, if it is to be serious, so far can not be included, for lack of appropriate properties. In itself this was not particularly urgent, as such names (usually) are not notable, and normally would not belong in Wikidata.
However, several Wikipedias do hold pages on such names, as a rule because of databasing run amock. An example is Albertia margaritifera, published in 1877 for a possible species in the family Umbelliferae. Since there already was a genus Albertia elsewhere, this attempt in naming was dead on arrival. By international agreement, of some two centuries standing (or a century and a half, if formalization is counted), such names do exist as names (that is, as separate formal entities), but cannot ever be used as the correct name of a taxon. In theory it is possible that later somebody can recognize a species in a different genus, and will find that this Albertia margaritifera represents the earliest attempt to name the species; in that case he would base a new name on this earlier attempt, leading to, say, Wikidatia margaritifera. However, this is not the case here and Albertia margaritifera was safely forgotten until databasing began.
The World Umbellifer Database recorded Albertia margaritifera, for completeness sake, as being a name safely forgotten. The Catalogue of Life copied the World Umbellifer Database wholesale, without bothering with pesky details, and made Albertia margaritifera an accepted species (that is, a species that exists only digitally, a database artifact). Then Lsjbot copied CoL into svwiki (and cebwiki and warwiki) (after all, if one wants to become the biggest Wikipedia, one should copy the biggest database one can find, so why worry about quality; this is not an isolated case). To be fair, the CoL corrected itself within a year, but Lsjbot did not follow this. So, there is a page on a fake species in svwiki, and therefore an item in Wikidata. It does not matter all that much how Wikidata handles this, provided it does not go along with this fakery and does not treat Albertia margaritifera as a real taxon (or Albertia margaritifera as a real taxon name).
Unfortunately, in early to mid 2019 a number of users decided to abandon the consensus model. Apparently, there are two groups of these: those who want items to focus only on names (ignoring taxa) and those who feel that if there is a scientific name (or something that looks like it), there also is a taxon. The effect is the same, Wikidata is now set to create many fake species (and other taxa), that exist only digitally. Thus negating the efforts of scientists who have laboured for two centuries to create order out of chaos: Wikidata is set to re-create chaos out of order. - Brya (talk) 04:37, 29 July 2019 (UTC)