The most interesting content in the survey consists in showing that philosophers are wrong or inaccurate in what they take to be the majority positions. Insofar as what the philosophers currently believe about what they believe is itself a majority position, the survey should have the result of changing a current state of affairs rather than entrenching it. (N.B. This comment shows that the accuracy of the survey pertains to your criticism because the survey should not be effective, if, for instance—and contrary to the fact portrayed in the survey—the majority is correct in believing what it takes the majority of philosophers to believe.) Since this result is both contrary to your position and very likely, your principle—that “at best, a survey of this kind can only encourage conformist, middle-of -the-road thinking.”—needs more support than gesturing at revolutionary progress. Not all progress is revolutionary, and some revolutionaries have been conformists—like Plato.

In every age, there does seem to be positions to which once must conform if one is not to jeopardize one’s livelihood, be ostracized, etc. Certainly, there have been marvelous, revolutionary individuals who have thrown caution to the winds. It is also possible that anonymous surveys taken in any age could have had the effect of showing that various “popular” beliefs were not so popular; hence, creating an atmosphere in which espousing “jeopardizing” beliefs is not contrary to self-interest. It’s possible that if such surveys were taken and made available, the mavericks would not have been “wild”, but geniuses arguing for positions akin to actual majority views. Not only are mavericks not always right, sometimes “independent thinker” is an oxymoron.  

(There are possible worlds where radical views are secretly, rationally, and widely held, and held silently until a survey is distributed and the radical private thoughts are made known. For something ironic, an instance would be the hypothetical situation in which a survey of enlightenment thinkers reveals that hard determinism is a very popular view. While this result would have been unlikely, this shows that some principle that this sort of surveys will at best entrench common views is likewise unlikely. It also shows to be false the principle that such surveys can only encourage middle-of-the-road thinking.)   

One can be wary of popularity without assuming that indication of popularity entrenches popularity.

Since you granted that the survey can be 100% accurate, let’s assume the implication: The views of the test subjects are represented with 100% accuracy. Since the test subjects were analytic philosophers, I am confused by your point about the unsuitability of the survey to represent the views of continental philosophers. What should one make of a person presuming to point out that a study using the expression “saying ouch” will entrench current theories about pain and inhibit progress because it is not reliable indicator of what people are talking about when they use the expression “awe, that feels good” ?  The inclination is to disregard. A study about pain is not about all sensations. An irrelevant statement: Continental philosophers are philosophers too, and since this survey is not about them, it will tend to underrepresent or overly restrict what philosophers believe. The study is not about all philosophers.