This is how the code is structured now (or should be).
First off, everything is an object. The Title as a whole is an object; it contains its own initialization functions. The same goes for Menus; they keep track of all their buttons internally. The buttons, in turn, know which sprites activate them, and which actions to perform for each type of cursor event. And the actions, like play audio, know everything they need to do their thing.
Each menu screen has an Initialization (Init) frame and a Wait frame.
In the Init frame script, a new Menu object is created, and buttons specific to that frame are created and added to that menu. Action objects (e.g. sprite visibility, play audio) are then added to those buttons as events (e.g. AddRollover, AddMouseDown).
In the Wait frame script, the exitFrame handler just loops on that frame. There are also handlers for mouseUp, mouseDown, and Idle. These handlers call the corresponding handlers in the Title object, which do some of their own things (like the Ad Banner) and then call the handlers in the Menu object. The Menu does some of its own stuff (like the Video Display) determines which sprite the mouse is over, and calls the handlers for its buttons. The buttons check for a match with their sprites, and call their actions.
One of the key points here is that the actions are independent of the event handlers. A Play Audio object can be created without regard for whether it will be triggered by a mouseDown, a rollover, or a timer event.
