That is complete conjecture.
If you and I, and anyone else can look at the results of that poll and think, "Yeah that seems congruent with my experience." Then there isn't much reason to question its methods and validity.
I have already given credence to the fact that there is room for error, and any statistician would note those potential errors when/if reporting such information.
It is still reasonable to use such information (especially when you have no other sources available to you) to have a discussion within this "hardcore" community.
Especially given the fact that the question the OP poses is presented to that... wait for it... hardcore community. The discussion that takes place directly refute that information or confirm it. If the discussion refutes that... the OP may get additional information to persuade him into participating in such content or reaffirm his/her decision not to.
I find it absolutely hilarious that this community thinks themselves qualified enough to predict launch dates, Blizzard bankruptcy, player engagement, performance, etc. based on ZERO relevant information. Yet when something like a poll is done (which is realistically just about the only semi-relevant data available to you, me, or anyone else) you guys want to sit here and pick it apart. Despite the fact that (while your points are valid) it probably reflects the community as a whole or pretty close to it, with a relatively reasonable margin of error.
You can accept information whilst having a reasonable amount of skepticism regarding its completeness or correctness.
Yes he/she probably came to the right conclusion, even if the methods to get there are wrong. It doesn't change the discussion.
Now if you have a better way to poll or collect information that is going to produce a significantly different result, that's worthwhile discussion. But even if you account for all of these errors that may skew the data, and the data comes out similar... we'll still be asking the same question. "Were Micro Holidays a complete failure?"