When I see "experiment X backs up/supports/proves conclusion Y" in a headline I always mentally correct it to "experiment X does not definitively disprove conclusion Y" where X and Y are random variables operating on whatever the actual scientific experiment and conclusion respectively are.