They're using state health hospital records which is a solid data set. But still looking at what is essentially census data, drawing out a correlation, and then making the leap to causation. They account for other demographic factors, but not what is arguably the other key factor which is cause of death.

Not bad science, just lazy. Definitely feels like a marketing thing where they began with a conclusion and used publicly available data to find the support.

All that being said, breastfeeding is definitely good, plenty of studies show this with more reasoning behind it. New parents should give it a try if willing and able. But the last thing new parents need is scare tactic statistics, choosing to formula feed shouldn't automatically have those parents lumped together with neglectful parents in statistics like these. Fed babies survive.