It is back to biased data.
When it was noted that people in certain occupations smoked more and got more cancer, tobacco was subject to a virtual jihad, denying any benefit from its consumption.
It has the opposite effects of alcohol. It is a stimulant, antidepressant, and constricts the peripheral blood vessels.
You see films of wounded soldiers smoking, even with torso (chest) wounds in WWII. Why? because it helped combat hypovolemic shock by redirecting blood flow to the organs by constricting peripheral vessels. It also offsets the effects of psychogenic shock to some extent (observation as an EMT).
But it is messy, does damage the lungs, etc. --all problems conspicuously absent from any discussion of marijuana.
I strongly suspect that cancer is a combination of factors, possibly pathogenic in part, some due to irritants (asbestos, smoke) and industrial and environmental exposures that got short shrift when tobacco became the boogeyman of choice. Genetics may play a greater role than suspected.
In short, those data have been compromised by questionable science, in that when one factor was identified, others were not looked for, and that abandonment of query has set back science through dogma. Something we are seeing yet again with COVID.