I'm not sure if this is poor science, or poor reporting. Some issues -
- The mentioned reduction looks like a total reduction, rather than the reduction attributable to cannabis. As a lot of the cannabis smokers in the study also smoked tobacco, that's an important 'oversight'. Smoking tobacco has been shown to damage male fertility in numerous studies - a 10 to 30 percent drop just from tobacco wouldn't be terribly surprising. They say they account for it, but it's highly unlikely we're being given that figure. From the dosage they were looking at I'd expect a range of 1 to 5 percent, probably closer to 1, though.
- Have they really accounted for tobacco? Surely they could have found some non-smoking cannabis users? Tobacco is a huge confounding factor. Furthermore, they don't seem to have checked general health levels - they even admit these factors could play a causal role. Seriously? These are things that can be controlled for.
It seems pretty flimsy to me. Without seeing the paper it's impossible to tell if they've found anything at all.
To be clear, I'm not saying cannabis doesn't affect male fertility, it may well do. I'm saying this is dubious evidence.