When an artist (or other entity for which images are displayed) doesn't have a link to Wikimedia Commons or to Wikidata, or when an existing Wikidata entry doesn't link to an image, JS code is put into the page to retrieve the ../commons-image endpoint in spite of this. The endpoint then loads the same entity data and relationships again and finds again that no image exists.

This is obviously redundant; fixing it may reduce the number of website requests noticeably.