The point here becomes that the original data source(s) are freely available and often unlicensed. Same example: the local bakery is not going to license their website information (or often the site is copyrighted by the creating company). You can copy this information to OSM. You can copy this information to Google Maps and OSM at the same time. Google cannot put this information under their license with a ‘no extraction’ rule because they don’t own that information. I get that OSM has to be cautious but no sane judge would ever rule in Google’s favor in this case.