I think it might work better in some subjects than others -- with the molecular pathology I read in my final year, we were expected to be completely conversant with current published research -- including stuff published after we'd been lectured on the topic -- so it's not clear how much more advanced it was possible to get without undertaking a more-prolonged or self-directed research project.

But in long-established subjects such as mathematics, a fourth year diploma was required as a prerequisite to starting a PhD, at least in my university, and I don't think even after the diploma students were up to speed on current research even across the topics they'd specialised in.