I agree with DHH and here is why ...

I’ve seen this pattern many times in my 55 years as a computer programmer. p(bloom) has always been greater than p(doom)