No one is claiming that this is a scientific poll. You'd have to be naive to assume that the results could be generalized to the population at large. That doesn't mean this poll is useless.
Yes it does. It's a poll of all the people who chose to do this poll.
I suspect that is massively biased towards people who did drop Dropbox.
At the very least the results useless unless we can estimate the self-selection bias.
Nobody but a fool would ever mistake a poll on HN for a poll of the general population.
The entire point of a poll on HN is to poll HN users, not the general population.
I would be amazed if even 20% knew about this in the first place, and amazed if as much as 1/10 of those people truly stopped using Dropbox.