According to a book I am reading the WTSD stat is accurate around 500 hands.
300 hands = poor
100 hands = aweful

With 3-bet stats, the book advices also a bigger sample size.

Most fish at 2nl, 5nl, 10nl, wont ever play 300 hands. Most fish, you play 1 session with.

Are stats like WTSD still useful?

I think it's a flaw in the book or a flaw in my understanding. The author states that a high WTSD with fish means we can valuebet wider. That's valueable informational, but then what do we do with fish who never reach a useful sample size?