Yes and no. Everything depends on the purpose of the survey. Your survey population must resemble as closely as possible your target population.
If the survey is intended to identify movies that would appeal to patrons aged 45 and older, it would not make sense to interview anyone else.
If it is intended as a measure of interest of the movie-going public at large, it is clearly biased, and given the demographics of the movie-going public, it would, in fact, be completely useless.
Another issue, a possible source of bias depending on the intent of the survey, is that the population of existing patrons aged 45 and older is less than perfect if your intent is to try to identify movies that would attract more patrons, who are not current customers, aged 45 and older. This is a common form of sample bias, because convenience equates with low cost. You end up significantly over-counting those who already like your product or service, simply because they are easier for you to find. (It is like looking for a lost coin under a streetlamp rather than across the street, in the bushes it rolled into, because the light is better.)
Your survey population must resemble as closely as possible your target population.
If you told me that you really, after careful thought, intended to sample this population, I would conclude your purpose was to see how you could improve your satisfaction/retention among existing customers in that age group. For any other purpose, it is clearly flawed.
So, yes, absent a clear, calm, and sober statement of purpose, and a solemn vow that you would never use the survey for any other purpose, I would say this sampling technique is unacceptable.
