How can we users contribute to better protect "crowd sourcing" projects like GBIF, OSM, Wikipedia and others against misuse?

You got my point.
It was easier before to defend data publication facing potential abuse because extracting, assessing, and bringing data into context was a substantial and labor intense effort.
This is changing now!
Maybe I’m too pessimistic, but I am much afraid that using GBIF for preservation will be soon less effective than abusing it for personal profit if the latter is supported by intelligent datamining services.