One of the least understandable things about en.Wikipedia is the low value of its own policies to its editors, in particular the low esteem for verifiability.
I am not the only editor who can see Cwmhiraeth’s howlers. However, the desired outcome of pointing out her mistakes is not that she or someone correct them, it is that I stop pointing them out.
This is like that nightmare geology article, declared a “Good Article” while containing time travel and imaginary rock formations. There is no reliable source on the planet for this fake science information, but, in spite of being entirely unverifable, the only thing that mattered was its promotion to “Good Article” status must be kept. Correcting the nonsense was not important. The best of Wikipedia is nonsense.
A “Featured Picture” contradicts the article text? It does not matter, because consensus, also known as majority voting, trumps verifiability.