Stockfish doesn't seem to be using the syzygy tablebases

[MichelinMan]

How to reproduce:

Below is example 6 piece end game, where the DTZ is 37
https://syzygy-tables.info/?fen=3qk3/8/8/8/8/8/8/2NQKN2_w_-_-_0_1

Load the same position into stockfish by copying the fen URL parameter into the clipboard and start and new game from the clipboard, then start infinite analysis.

Expected outcome is to get the forced mate result immediately.
Instead Stockfish analyses the position conventionally, and 5 mins later it has a mate in 23 on its own, not consistent at all with using the TBs

Note:
My TB folder configuration is correct.
All 730 6 piece TB files has their checksums validated and are all correct.
I don't know what/how to troubleshoot this.

Happy to help with any questions.
Thank you

[MichelinMan]

I see this comment in the docs, so perhaps I am seeing the expected behaviour?
It is very strange that it cannot find the optimal play even after 5 mins, strange to me at least, but I am not an expert on this subject.

https://official-stockfish.github.io/docs/stockfish-wiki/Advanced-topics.html#syzygy-tablebases