That's where the timeliness versus accuracy trade-off comes in. There is significant value in having an early signal to act on, especially as long as there is awareness of its limitations. And as I mentioned, this data does come with very clearly documented caveats and methodologies so that users can make informed decisions.

Normally they're accurate enough, and the revisions/refinements are small.

The initial estimates haven't been accurate recently, because many workers have been completely dropping out of the economy because they're afraid of immigration raids. Information about these workers is just harder to gather and takes longer to verify.

Unless they are accurate most of the time, I would agree with this.