To go along with what Andy mentioned , a few things I'm personally reminded of following this event:
1. Watch for situations where bulk shear vectors don't back behind initiating surface(s) - katafront. Outbreaks don't always work in this manner, however it certainly adds an additional layer of confidence when observed. DPVA is of a special kind when this occurs.
2. The presence of storm relative helicity is strictly a necessary yet insufficient item within the tornadogenesis process. Often seemingly meaningless details in the storm relative wind profile throughout the depth of the cloud bearing layer are, and always will be, king. However, the wind profile entity is what controls the storm relative flow via storm motion, and thus the ground relative/storm relative puzzle results. Buoyancy acts as clay and wind shear the potter, so many possible resultant shapes and sizes.
3. Ingredients-based forecasting is crucial in our field. It absolutely does not matter if there has never been a tornado outbreak associated with a surface low in lower-middle Alabama, if the atmospheric parameters indicate tornadoes, then a tornado forecast it is.
4. The exit region of upper jets in general is an excellent place for tornado outbreaks. Most times, it is best not to place emphasis on jet quadranting for many reasons, but rather on the presence of diffluence which overpowers speed convergence - which in today's case consisted of a nosing upper jet over top the western edge of the warm sector air mass.
We study a fascinating, powerful, and humbling science, and I just hate that we lost people today in this outbreak...