Yes, I got that impression too. That's why I'm saying any such conclusion is without merit, done in the fashion it was. Not to say the idea is a bad one, but a real poll would be expensive and involve much research. A random sampling of a cross section of all socio-economic levels in each country would need to be made (minimum of 30 people in each grouping.) This could be done by mail if you had the addresses. Or go knocking on doors, which would be risky and possibly illegal in some countries. May be hard to get permits to do that as well, especially considering the nature of the poll.
If it took too long, the poll would time itself out and the results would be full of errors. The questions asked on the poll should be checked ahead of time by a qualified linguist or two, to make certain they are not skewed. Even the order in which questions are presented could skew the results.
There would probably be a large margin of error due to mistrust on behalf of people simply not answering the poll.
In my opinion, conducting such a survey would be a Herculean effort if not impossible. However, results from such would be scientific and accurate, for whoever had the fortitude to do it.
Les