That's great and all, but nobody needs a 32-bit anything in 2018. This undergraduate paper provides a magic number and associated error bound for 64-bit doubles:
-3 karma · joined January 31, 2014
No one in the public discourse is saying our thinking on vaccinations is being continuously refined. No one is publicly urging caution and moderation in evaluating the risks and benefits, or calling for better longitudinal studies. Instead, on the one had, you have the white house saying "the science is clear," people calling anti-vaxers morons on social media, and on the other side people yelling that its poison pushed by a global conspiracy. Whatever calm, analytic discussions are being held behind closed doors at the NIH, the public discourse has degenerated into a screaming match.
Adam's problem with that is that the pro-science people are screaming just as loudly as their opponents... but they should be the ones that know better! Unfortunately, people of a certain age have seen dogmatic claims from "the scientific community" before, and then seen those claims be overturned and reversed countless times. That's a problem. If you really are "pro-science" you should ALWAYS be couching your recommendations in uncertain terms, but there are lots of demagogues out there doing exactly the opposite.