But I really care about "there's a 95% chance the benefit is somewhere between 1.9x and 5x, so it's worth it."

So that's why CIs are interesting and p-values are less so.