I don't see any part where the talk about the "amount" of drugs consumed?
There were rougly 1600+ babies who were used as control and ~338 who tested positive for "weed". Some of the things that aren't broken down is the distribution for infant mortality. Perhaps there were outliers in the control group such as premature babies, which, if taken into account might explain some of the differences in mortality rates.
Very funny though, with actual data to back up some part of your claim, made me think... :)