I didn't conduct the survey, but typically a sample size of 500 is considered sufficient... so 75,000 is WAY more than needed for an accurate depiction of the situation (150 times the sample size of your average poll, scientific study, etc)
I would agree. But only partially. A sample size of 500 is enough, if it gives the necessary weight to all the major demographic groups.
For example, a sample survey about the presidential race will consult 500 voters, out of whom approx. 350 will be non-Hispanic white, 75 African-American, 50 Hispanic, and 25 Asian and others. Gender-wise, they will have equal representation from males and females.
Are you sure that the surveyors have given accurate weight to the major groups here? For example, if the survey had an over-representation from New Mexico or California, it will have a bias towards Catholics (due to the Hispanic population). On the other hand, if there is an over-representation from the south-east, there will be a bias towards protestants.
That said, I don't claim that this survey is an outlier. The results, match with my expectations.
Look at any comparable statistic, and you will see the truth... don't take my word for it, do your own research
They originate from a .gov website mentioned in the image (Bureau of Justice Statistics - Federal Bureau of Prisons)
Feel free to read all about their data collection policies if you are interested in such things
On the topic of Buddhism...
I appreciate that Buddhists do their own research and do not hassle me about something they could easily google themselves