Real economists and economic observers have been following this. I will just add observations from a purely engineering/math perspective.
First the kindness. at first it looks like the two metrics are irreconcilable. Not as bad as it looks. GDPNow from Atlanta is subject to wild swings being a snapshot. The NY Fed evolves. Trading sensitivity for integration is merely choice, not bad methodology.
Now the criticism. There is no way both are correct. There is no overlap even allowing for revisions. Do you think come the release of the accepted GDP that one or both will admit to being wrong?
Worse, if one is lucky then we tend to dismiss the other. We need a lot more data points.
Personal note. More weeding. Pulled more invasive grass and discovered this patch of teacup roses and mixed ivy plants.