Re: I don't think so...
Oh lordy, I do love going through the stats on any of those "pot causes x" papers. Becasue as you rightly point out, either their math is waaay off, or the correlations are pretty weak. And never any causation either :)
The number of psych and social studies students who just don't seem to get statistics is mind boggling. All the "science" in these areas is based of data that is statistical in nature. The theories and philosophies are all nice, but there seems to be a strong desire to make the data fit the theory, rather than devolop theories to explain the data.
My favorite from a few years back was the headline grabbing "Pot makes you 3 times as likely to get schizophrenia" which was usually expressed as 300% increase in risk (since journo's hate stats even more) when the actual data was that schizophrenia in the general population was 0.9 per 100k, heavy pot smoking and schizophrenia was 2.6 per 100k. All well and good, apart from the small fact that the same study found that alchol abuse and schizophrenia where roughly 1000 per 100k, and heavy tobacco use and schizophrenia roughly 2500 per 100k. No mention of smoking or drinking being much more likely to "cause" schizophrenia*.
TL:DR - Study shows weak correlation between mental illness and pot use, strong correlation between alchohol or tobacco use and mental illness. Media only reports on pot aspect.
* Should also be noted that the study as published was only looking at correlations, and the causation attributions seems to only have been added by non scientists. Correlation is not causation, unless the result backs up your worldview :)