Actually, just as a point of order... it is not necessary to look at millions of passengers to determine "the facts." Statistically:

If you have a representative sample and use a sample size of approximately 1000 randomly selected respondents, you will have a sampling error of +/- 3 percentage points at the 95 percent confidence level. ("Sampling error" is the variability that occurs because only a small portion of the population is studied, rather than the entire population. ie. looking at some people at airports instead of every single person at every airport).

What does this mean? It means that if, say, 37% of travelers experience something, then you can be certain that if you kept doing the survey over and over and over again indefinitely, the results would be that between 34% and 40% of travelers would have that experience in 95% of your surveys.

By the way, just in case you are wondering where I got this information... it is a general principle of statistics. You could look at the Editor's Note page of American Demographics magazine if you want to see a reprint of my explanation. (although they don't mention shoes or airports... )

So, if you're really curious, assume that the first 1000 people who walk through security are "representative." ie. don't count crew, TSA, FAMs, etc in the 1000 -- just use "regular" passengers. See what happens.