Honestly, without knowing the exact question it's hard to predict what the numbers will be. For example, there used to be a ~15% spread between people who supported gays and lesbians in the military vs. homosexuals in the military. [With another ~15% gap between 'serve in the military' vs. 'serve openly in the military'. The things we choose to care about...]
There is almost certainly a question phrasing that would yield something similar to your general sense of people's opinions (once you adjust for the fact you do not interact with randomly selected people, and tend to discuss the issue only with those who feel safe sharing their opinion with you).
There is almost certainly a question phrasing that would yield something similar to your general sense of people's opinions (once you adjust for the fact you do not interact with randomly selected people, and tend to discuss the issue only with those who feel safe sharing their opinion with you).