On the other hand, the textual information provided is sparse and occasionally *wrong*, and some of the inclusions are questionable.
Each illustration has a one-sentence caption, varying considerably in detail provided. Some give you all the information you would need to find the source of the original illustration, some are a single word identifying the type of creature, others are somewhere in-between. It's unclear how much of this dearth of information is because of the reliance on secondary rather than original sources, and how much of it is because the author and/or his editor over-did the cutting.
Also, some of the captions are mis-assigned. Going through the book, I easily spotted several occasions where two of the illustrations on the same page were clearly given each other's captions. I only spotted this a few times but it begs the question, how many other such errors are there in the book that I didn't spot?
Finally, some of the illustrations included really don't belong. Details from Bosch paintings make sense, they're demons ("fantastical and mythological creatures") from hundreds of years ago ("from historic sources"). But there are some figures from modern art included that don't seem to belong, particularly a Picasso that's just an abstract rendering of a normal human.
It may well be more useful for its bibliography than for its content.