Network of people who have access to the target population sample is one way of how those in the academics get their thousands of participants. And basically, they do have their research assistants to collect their data for them. Web-based surveys may be helpful but reliability of responses remains to be confirmed. Some use googledocs but they still need a network to gather responses, usually through snowball sampling.
I’m not sure how to put it, but it’s a question of higher response rate vs accuracy, reliability and validity of responses.