I think a clear distinction needs to be made between normative and descriptive claims: the former, everyone has the right to make and hold their own opinion to,

I think a clear distinction needs to be made between normative and descriptive claims: the former, everyone has the right to make and hold their own opinion to, including both politicians and scientists; the latter, should only be entertained by policy-makers when supported by evidence-based scientific research. For example, a politician has every right to say "I think schools should re-open immediately under X procedures," if that is what he truly feels reflects the best outcome for his constituents. Of course, it would be wrong to say this only for political posturing, corrupt influences, etc. On the other hand, it is completely unacceptable for a politician (or anyone) to say, "Children are immune to the virus," "COVID is not a serious illness," (okay, technically that is normative but only insofar as the definition of seriousness is normative), "same as the flu," "masks cause oxygen lack," etc. Likewise, asserting these claims in an argument, "Children are immune -> we should open schools," is not right. But, making a philosophical/policy and/or medical argument based off true, evidence-backed descriptive claims--even if the scientific community does not tend to share that normative opinion, is completely fine. So there is nothing wrong with arguing against a given measure that the majority of doctors support, so long as you aren't distorting the underlying facts. As a practical matter, of course, this shuts down many arguments. Ex. admitting that masks pose no health risk to the wearer, cost very little, do not substantially limit any right or activity, and have at least a fair probability of public health benefit, it would be hard to argue against requiring them. But if you feel otherwise, make your case by all means. The only thing that is morally wrong is corrupting intellectual discourse by adding FUD and confusing people as to the underlying evidence and data.