Partitioned/Incremental dataset: To load or not to load?

Hello everyone!

The other day, I had to switch the input of my function that concatenates data from an incremental dataset to a partitioned dataset. I thought this would require just modifying the data catalog, but it turns out you first need to load the items in the dictionary of your partitioned dataset. An example is shown in this video @3.51s where csv = load_csv() is used. My question is why this is? Does anyone know if this was a design choice or just a coincidence? When I look at the source code, the load functions look quite similar with the main difference being the parentheses:
Partitioned: partitions[partition_id] = dataset.load
Incremental: partitions[partition_id] = self._dataset_type( # type: ignore **kwargs ).load()
https://kedro.readthedocs.io/en/stable/_modules/kedro/io/partitioned_data_set.html#PartitionedDataSet
https://kedro.readthedocs.io/en/stable/_modules/kedro/io/partitioned_data_set.html#IncrementalDataSet

Cheers!