Yeah, I agree with Felicity
I'm pretty sure the categorical approach has higher inter-rater reliability. If a mental health professional used the dimensional approach, most of the time they'd have to create their own questionnaires or standardised inventories to assess a particular psychological characteristic. If all these mental health professionals have different ways of assessing the same characteristic, there's not going to be a great deal of consistency
With the categorical approach, they're really all using the same diagnostic criteria (which is outlined in the DSM and ICD), so there's going to be a lot more consistency in their diagnoses. Of course, the symptoms for some distinct disorders are going to overlap in a pretty big way (especially personality disorders). But on the whole, diagnosis is going be more consistent than if a dimensional approach were used