Modifiers allow you to create a separate action using the same (or no) gesture only differentiated by an extra event, like pressing the Control key. This would allow you have the same gesture do different things based on what modifier was also present.
For example, maybe you have a action that minimizes a window using the straight down gesture. But you have another action using the same gesture but uses the Control modifier, so when you draw the gesture and press Control it actually closes the window instead of minimizing it. Of course, the actions don't have to be related in any way at all, but it is easier to remember.
You can choose to recognize modifiers:
Essentially this allows you to be even more specific about when the modifier event occurred. Before means it would only match if the modifier was active at the time the stroke button was pressed. After only matches if the modifier was active after the stroke button was pressed. Either means you don't care, just as long as it happened.
The reason for this distinction is to support granular control, but is also helpful when creating action pairs like rocker gestures.