Hi there! Thank you for the praise! Let me break down your questions:

1. It is possible to generate these statistics and to figure out the biodiversity and proportion of endangered species sightings in comparison to the other species present. The two biggest issues are uncertainty of location and the size of the data in question. Because these animals are protected, some of the data is obscured to decrease the likelihood of poaching. This makes certain spatial analysis tricky and can lead to a decrease in accuracy. Second is that there are billions of sightings logged in GBIF and comparing these together even at smaller scales like a country or state become challenging. My lowly MacBook Pro and free GIS software simply doesn't have the bandwidth to conduct that level of analysis (although I really want to!). All of this to say, the solutions you've presented are possible!
2. The way a species is defined as endangered for this map is controlled by an international agency. While different states and countries can create their own endangered species list, IUCN is the most internationally recognized. This eliminates bias on the local/state level but doesn't remove our charismatic animal-centric bias which is a big problem facing conservation.

Hopefully those answered your questions!