The interesting thing is that you can't actually divide one number by the other, because you don't know when the marriages began. What you actually need is the data "how many marriages in 1980/1981 &c end in divorce"? The information that a bunch of retired baby boomers are divorcing has little to no relevance to 30-somethings getting married right now.

In particular, since fewer people are getting married than used to (the whole point of the article), this is overstating the divorce rate.