Part of the problem is the faster and wider communication of these kinds of incidents. I think people can intuitively understand, at lowest, 5%* odds. Anything below that is either 5% or impossible unless you're mathing it out.
Hearing about these kinds of tail-case incidents at least once or twice a year is probably enough to bump them up to the 5% category.
*I'm not committed to any specific number here. The relevant threshold is "high enough to consider when making decisions".
An interesting resource is IDC 203, which describes how the US intelligence community discusses probability. The table in paragraph 2a says something similar to what you are saying:
almost no chance / remote := 1%-5%
very unlikely / highly improbable := 5%-20%
etc...
Hearing about these kinds of tail-case incidents at least once or twice a year is probably enough to bump them up to the 5% category.
*I'm not committed to any specific number here. The relevant threshold is "high enough to consider when making decisions".