perf: remove unnecessary qs distinct call

When filtering using Exists() there is no risk of introducing duplicate
rows.

http://localhost:8000/v1/opening_hours/?data_source=tprek&start_date=0w
&end_date=4w&page=3&page_size=10

Query time goes from 30-40 ms to <20 ms

refs: HAUKI-647

hours/viewsets.py

@@ -989,7 +989,6 @@ def get_queryset(self, start_date=None):
                     ),
                 ),
             )
-            .distinct()
             .order_by("id")
         )