Well yes, but how many times does that actually happen?
Events have to be separated and categorized. You have to have one category designated to your classic, textbook setups and events and one category that isn't classic or textbook. When you have your classic, textbook setups you know what's going to happen...and that is true for anything (winter, severe, tropical).
However, how many times do we actually get those classic, textbook setups? Not very often. So analyzing the non-textbook setups and trying to find as many different discriminators as possible and determining which of those discriminators holds significant weight (which is going to differ by setup) can be quite telling. It's absolutely very tedious work, but IMO understanding this would significantly improve forecasting at least in the medium-range (3-6 days out).