Survey results always have to be taken with a grain of salt. The inevitable flaw of surveys is that they merely measure a small subset of the total population, namely that of "people willing to participate in surveys".
Also, cross country comparisons are going to have issues no matter the format. Data collection methods will inevitably vary between countries even beyond rigor and completeness. You also have to consider the language translation issues.
I'm not saying that this study is wrong or that the problem doesn't exist. I'm just saying you have to be very cautious drawing conclusions from these types of data.