Well, there are several reasons why I'm not incessantly shouting all my contrarian views from the rooftops.

For start, yes, obviously I am concerned with the possible reputational consequences. But even ignoring that, the problem is that arguing for contrarian views may well have the effect of making them even more disreputable and strengthening the mainstream consensus, if it's done in a way that signals low status, eccentricity, immorality, etc., or otherwise enables the mainstream advocates to score a rhetorical victory in the ensuing debate (regardless of the substance of the arguments). Thus, even judging purely by how much you're likely to move people's opinions closer or further from the truth, you should avoid arguing for contrarian views unless the situation seems especially favorable, in the sense that you'll be able to present your case competently and in front of a suitable audience.

Moreover, there is always the problem of whether you can trust your own contrarian opinions. After all, even if you take the least favorable view of the respectable opinion and the academic mainstream, it is still the case that most contrarians are deluded in even crazier ways. So how do you know that you haven't in fact become a crackpot yourself? This is why rather than making a piecemeal catalog of delusional mainstream views, I would prefer to have a more general framework for estimating how reliable the mainstream opinion is likely to be on a particular subject given various factors and circumstances, and what general social, economic, political, and other mechanisms have practical influence in this regard. Effort spent on obtaining such insight is, in my opinion, far more useful than attacking seemingly wrong mainstream opinions one by one.

These latter questions should, in my opinion, be very high (if not on the top) of the list of priorities of people who are concerned with overcoming bias and increasing their rationality and the accuracy of their beliefs, and one of my major disappointments with LW is that attempts to open discussion about these matters invariably fall flat. (This despite the fact that such discussions could be productive even without opening any especially dangerous and charged topics, and despite the fact that on LW one regularly hears frustrated accounts of the mainstream being impervious to argument on topics such as existential risk or cryonics. I find it especially puzzling that smart people who are concerned about the latter have no interest in investigating the underlying more general and systematic problems.)