Density-functional theory and experiment: a match made in heaven or in hell?