I wonder what an optimal portfolio of anti-existential risk programs would look like

Agreed that it's not obvious, and what's more, it would almost certainly be different than what one would expect to be reflected in the tag cloud of LW posts about what an optimal portfolio of anti-existential risk programs would look like.

E.g. global warming is an important issue, but it's not important for LW to dilute its content by being the millionth website to feature content on it.