I see your point. It probably depends on what is the reason for the hackathon.

I've seen couple of "private" hackathons where companies shared their data (anonymised and with NDAs) and often (semi-)professional data scientists were using them to solve predefined business problems.

The extra benefit for the company was the fresh data solutions that "outsiders" saw in the company's data.

Data scientist with corporate, consulting and start-up experience | avid cyclist | amateur pianist | CEO & co-founder at DataDiligence.com