Your hypothetical examples are interesting. Since you explicitly don’t care at all about the accuracy of the survey results, let’s assume that the hypothetical surveys taken in their respective centuries are completely accurate. Since you cannot get an ought from an is, the purely factual information would not be normative. Hence, I don’t see how getting clear views of what philosophical views were most widely believed implies that some negative phenomena would have become further entrenched; the actual views may well have been less ridiculous than those submitted to by public persona. If you assume that they might have lied and the surveys would have entrenched the persona opinions, then your worry is about validity—anonymity is assured so surveys are not defeated by this type of confounding variable: the notion subjects might have that they have to be politically correct. Moreover, I’m sure historians of philosophy would be overjoyed if they could actually get their hands on such data, rather than relying on texts of questionable accuracy.  

You worry does seem to be that the survey does not accurately delineate areas, problems, and views in philosophy, and acceptance of the survey implicates a sterile and facile notion that the constantly moving targets, philosophical positions, etc. can be captured in pat stereotypes. Further, a survey of this sort is likely to have the adverse effect of entrenching these stereotypes. If this is true, then explanation of the validity of the survey is appropriate. Just as correlations involving “saying ouch” can give useful information about pain, so too can correlations using rough demarcation strategies give useful information about what philosophers believe.

Your charge of complacency seems at least partially motivated by supposing that philosophy is inaccurately represented and that complacency is responsible for this inaccuracy and will entrench it further if allowed to. If the questions and answers do not reflect what philosophers believe (philosophy is not accurately represented), then the survey is invalidated. Hence, my defense of the validity defeats the charge of inaccuracy—except, perhaps, the charge that the aesthetic questions and answers inaccurately represent. This should mitigate the force of your complacency charge.   

On entrenching stereotypes:

 If the survey is accurate, what is the complacency? Your view is that complacency is built into the survey. If this doesn’t affect the validity of the survey, then it is far from clear that your complaint has any more force than saying “Boo” at the survey. After all, if the survey is valid and the testers and subjects are complacent, then what philosophers believe is complacent. There is no a priori reason to think that the data representing their beliefs justifies them in anyway. Ad populum is a fallacy. I’m sure the community is not as complacent as to trade in fallacies. (If appeal to expertise is appropriate, I can see how “complacency” might become entrenched.)

I fail to see the force of your complacency charge if it is not based on assumptions about the validity of the survey. A reasonable complaint is that the survey doesn’t measure what it’s supposed to measure; namely, what philosophers actually believe. IF it is true that you are not remotely interested in the accuracy of the survey, this should not be your worry. If you mean to merely point out that philosophers are complacent, then not only does your complaint seem not to have much to do specifically with this survey, but it also seems apt to be defeated by pointing at work done in most any of the top philosophy journals.

Nate