While there is nothing necessary moronic about the question - there are questions about compiling the data. The challenge is what data stream you are analysing and how that is really impacted by other factors.
To wit, many people in Houston know girls at various strip clubs who work on the side, and thus they can bypass the usual ECCIE and P411, etc. Some people use Eros, some people go for BP girls.
Personally, if I know a girl who works at a specific strip club and I know her rates, her phone number, etc. - why bother with the need to find a new provider, go through some meaningless and trivial "verification" and make a slew of arrangements when I can just call Jasmine up directly when I know she isn't working and she's dropped the kids off at school and have someone familiar who knows what I like and thus maximise the time?
That all being said, how do you measure the above in your data set to say why there are not so many online? Clearly, the online numbers do not reflect the reality. A friend of mine, a woman who made a PhD career on researching why women go into this line of work and the troubles they had in their childhood, noted that a lot of women would never get on sites like these and stay very much under the radar. They are cocktail waitresses or dance at strip clubs, they double down with jobs at Hooters or Twin Peaks, they make their money and represent the large contingent of women working in the industry who do this sort of thing and you never see an ad anywhere.
Cities like Houston have huge numbers of these women. And there is no racial component to it - thus, it is not about Latina Agencies or any such things either. You also miss out on numbers with local My Free Cams girls who live in the Houston area and whom, if you befriend online and build a trust bridge, will gladly give you a "private show" and you have no way to Google search those.
When you add in all the myriad of women who do the work under the radar and never promoted in Houston, the number shoots up far beyond the online count.
Cool .. Yes... This research is all based on the fact that
TAMU did a study on putting in golf....
The result of the study: 50% of all putts
Go in the hole..