@bampo 
Partitioning is splitting the data in multiple parquet files, that's why it's reading the table partially.
Liquid clustering behaves differently. With that small amount of data, it's keeping all of the data in the single file, because it's the most efficient way of handling a small data volume.

This video explains what liquid clustering is about - https://www.youtube.com/watch?v=5t6wX28JC_M