Sorry for jumping in so late. I absolutely agree with the points made here. What a great conversation, indeed!
I can imagine that useful quantitative information will be hard to find. At Visit Flanders, when doing a quality-oriented survey, we usually add a question: “Do you, or any of your travel companions, experience impairments or disabilities when travelling?” It’s a way to detect a part of the accessibility requiring segment by self-identification. Usually, it allows to make significant comparisons in travel behaviour and appreciation between the group that replied positive to the question and the entire population. But we don’t learn anything about those who have accessibility issues, but don’t acknowledge them in the survey, neither about those who would identify as disabled, but not when travelling.