Then the headline is misleading. This ignores gay men who don't go to bars, dance clubs, etc. and it ignores gay men who live outside the 21 AIDS hot spots. It would be like assessing Democrats by sampling only Berkeley and The Village. Conclusions about a self-selecting sample should be properly qualified. Like those car insurance adds about how much money people saved by switching "on average". Well, who is going to switch to a company that charges more?
For example, one of my daughter's friends and one of my high school classmates have long-term relationships with other men and they go to work each day and come home like the rest of us. They don't go out "clubbing" looking for sex. They would be missed by this sampling even though they live in NYC, which is likely in the top 21.
Headline writers are traditionally the lightweights in any newsroom. Some of them were likely part of this study...
You anecdotal experience is so rare as to be statistically insignificant and it does not alter the validity of the results for the vast majority of that 1.75% of the population.