Consider removing partitioned variants of Parquet, Data Lake, Vortex

[alexey-milovidov]

They are almost identical to "single" variants, and every reasonable entrant should give about the same results.
These almost identical results clutter the report.

The files will remain available for data loading.

[KazukiKandaKK]

I can take a look at this and compare the single and partitioned variants. If they are effectively the same, I can put together a cleanup PR.