Yeah, this makes me think further and it feels like one solution could be legislation that declares companies never own user-supplied content -- that when users submit data like Amazon reviews or Craigslist posts they become public domain. Competitors are free to scrape them however they can.

It's harder for map corrections though, as the user-presented data mixes commercial and user-supplied data. Maybe legislation should require regular data dumps of all user-supplied content much like Wikipedia makes available in XML form? Then no scraping is even required.