Although I agree, in principal, that having everybody over 55 lumped into a >55 category is not ideal, as a statistician who does data analysis for cancer clinical trials, I have to point out that with only 4 folks total out of the 135 respondents (at the time I'm writing this), you aren't going to gain that much more knowledge
1. It depends on the purpose of the poll. Do you want to try to gain an understanding of the full range of who enjoys biking and also participates in this forum? If so, you are losing valuable information - such as that, perhaps, someone 80 years old is participating in biking and is a forum member. Now, this is not statistically significant, but it is significant in learning about our membership and about the fact that you can be 80 (if that might be the case) and still enjoy biking, and, perhaps, the change in attitude of those who think they are old at 40 or even 50.
On my first Ride the Rockies, there was one person who was 83 who rode the whole thing. Statistically insignificant, yet this person made the TV news, while many in their 30's did not. This person would not have shown up on this poll.
2. You do not know how many, such as I, are refusing to participate in this poll because of the way the age categories are set.
3. And you have no legitimate knowledge beforehand that there will be only four respondents over age 55. That is prejudicing the survey before it is even given. You would not give a cancer survey leaving out a whole area of possible causes, for example, and then, post-survey, state, "Well, we had only a few responses in that area, so it was okay to structure the survey that way." At least I hope you would not.