I don't trust the GFS/FV3 operational look. It goes strongly against its own ensembles, shows very low run-to-run consistency, and it's obvious that the FV3 is suffering from feedback, with it focusing on an MCS-driven surface low in south AL as the primary low, only to magically jump to south central KY with the primary low 6 hours later. The vast majority of ensembles and non-GFS/FV3 operational models have an unsettling surface low track from Arkansas into Kentucky/Indiana and a low-amplitude negative tilt trough blowing through here. I'm still not ready to fully jump on it yet, but we're rapidly running out of time for this to just be model noise...