Sunscreen is a very poor example, if you’re trying to get any scientific research. There are a vast amount of variables that would contribute to cancer rates.
For instance, sunscreen is not measured, nor applied in any quantifiable amount. It is not applied evenly by some, and over applied by others. So. You can’t verify that most users of sunscreen even used it as directed. And there are way to many variables in the amount of radiation people are exposed to, due to the Sun (think time of year, altitude, cloud cover etc.)
Studies involving 500 people who used sunscreen, could have wildly different results. Not like using an appropriate mg/pound dosage of a drug on 500 people who are trying the medicine for a similar condition.
Lots of people in recent decades have gotten skin cancer on their forearms, and hands. My understanding is that the consensus is that this is a direct result of people having those areas exposed to the Sun while in automobiles. Truck drivers get it on their left arms a lot, in countries where they drive in the left side of the vehicle.
If you count natural Sun aid, like Aloe, or even mud, then sunscreen predates automobiles by thousands of years, with only about 50 years or so of having a tenuous grasp of how populations whose exposure to the Sun is affected by driving automobiles.
Scientific researchers often stay away from results rendered from such chaotic, random tests.
Science prefers as much control over an experiment as is possible. Even then, it is open to scrutiny by others. And, of course, the results must be able to be reproduced by others.
If one researches Terraforming, then you have the information required to prove that global warming is worsened by greenhouse gas emissions. The science is over, and accepted by (last I heard ) over 95% of climate scientists. Such a high rate of agreement is as close to 100% as you will find in any scientific research…