'generally safe' implies that some small number of folks still suffer from excess iodine due to contamination, manufacturing errors, etc...
It's probably reasonable for 999 999 people to benefit in exchange for 1 person being very unlucky, but that's a different argument that needs to be made.
Which is nothing? There's a clearly non-zero rate of dangerous contamination in food products, in a market as large as the US, that nobody is able to fully prevent.
Like I mentioned to the other replier, a probabilistic argument is fine, and I would even agree it's reasonable for 999 999 people to benefit in exchange for 1 person being very unlucky, but that's a different argument that needs to be made.
And nobody, in this post at least, has made that argument, nor has anyone even linked to such.
e.g. manufacturing error such that a dangerous amount of iodine is accidentally added.