Functionality implemented in API. Some considerations:
STAC library is slow, and the current implementation requires saving and loading the full catalog multiple times. Works fine for small datasets, but I am afraid this will not scale.
Only spatial duplicates are computed, but implementation of additional metrics is straightforward.
It would be nice to test the implementation and performance in a diverse set of STAC datasets.
Implement automatic QA for Q2+ datasets in API (already available in library).