Instead of presenting the consumers of gestures with all the details about the gesture we just give them the action that should be triggered, and whether we should go forward (e.g. switch to the left with workspace switches or open the multitasking view) or backward (the opposite).

This helps

1. with memory since we have to store less values (negligible but nice to have ig :)))
2. with performance because we don't have to call GestureSettings.get_action twice in e.g. WindowManager
3. with lines of code
4. with preparation for more deduplication between one to one gestures and e.g. keyboard "gestures"

So win, win, win, win I guess? Let me know what you think :)