Or, how I spent my morning.
The Herzog-August-Bibliothek in Wolfenbüttel just released a new Paracelsus facsimile. Many of Paracelsus's works were printed in the late 16th century, but this was a copy of Paracelsus's Prognostication for 24 Years, one of the most intriguing treatments of prophecy in word and image in the 16th century (which therefore plays a starring role in Chapter 4 of Printing and Prophecy). Heinrich Steiner printed two Latin and one German editions in Augsburg in 1536. According to the HAB, their Latin edition was also from 1536, so I assumed I would find another facsimile of one of Steiner's Latin editions.
I didn't. The HAB copy, place and publisher unknown, has an entirely different title formulation and text layout, and copper etchings rather than woodcuts. I couldn't find any trace of this edition in VD16. Could this possibly be a third Latin edition that had escaped my notice?
Alas, no. A quick look at the Karlsruher Virtueller Katalog found two more copies in Germany (BSB München, SBPK Berlin) and one in Austria (UB Salzburg), and Salzburg and Munich date this edition as ca. 1580. That would make much more sense, considering the quality of the images and the boom in publication of Paracelsus's works after 1560.
Extending the search in the KVK outside of German libraries turns up even more copies: University College London and the Wellcome Library in the UK, Toulouse (with a PDF facsimile) and six more copies in France. WorldCat finds 11 copies in U.S. libraries (but I'd treat that figure with some caution, as some American libraries don't always distinguish between originals and facsimiles in their catalogs). This isn't a distribution of copies that points strongly to Germany as the place of origin, by the way.
The record from the Wellcome Library catalog provides to the answer: "Referenced by: Sudhoff, K. Bibliographia Paracelsica 504." The fundamental bibliographic work on the editions of Paracelsus is still Karl Sudhoff's Versuch einer Kritik der echtheit der Paracelsischen Schriften (1894). For this edition, Sudhoff ignores the date of 1536 on the title page and treats this edition as undated. Sudhoff notes that the duration of the prognostication is twice changed from 24 to 44 years and suspects that it was printed, perhaps in France, around 1580 (=1536 + 44). The relatively high number of copies in France suggest that Sudhoff was on the right track.
As usual, I've needed a few hours and access to several different computer databases and digital facsimiles to get up to speed on the state of the art from 117 years ago.