>findings from a limited number of well-designed epidemiological studies do not suggest an increased risk for the development of either lung or upper airway cancer from light or moderate use
This is not a decrease of lung cancer - but merely not an increase. Meaning it shows it does not cause cancer like smoking tobacco.
To continue on...
>although evidence is mixed concerning possible carcinogenic risks of heavy, long-term use
Which means results are inconclusive for heavy, long-term use, meaning marijuana could still increase the risk of lung cancer by introducing carcinogens.
>In summary, the accumulated weight of evidence implies far lower risks for pulmonary complications of even regular heavy use of marijuana compared with the grave pulmonary consequences of tobacco.
The purpose of this was to show that even heavy use of marijuana has far less risks (but still has risks over not smoking at all) when compared to tobacco.