Right… that’s what I’m calling the “two-layer solution.” It definitely works, and I do it when I need to, but avoid it because it means more geometry to create, more to edit when things change, and more layers to manage properly in however many scenes I need. And if (as is often the case) I want to just show a particular cabinet with doors open and the rest closed, and maybe a different cabinet open in another scene, it gets very complex very quickly.
That’s what would be nice about the DC solution. One copy of the geometry, with granular control over each component.
The DC solution is not a current solution for LayOut, because LayOut does not have a Ruby API, and may never get one. Secondly, it is the Ruby code running inside SketchUp, that animates the normally closed cabinet door to swing open. This can only be triggered inside SketchUp by a manual click of the DC (or some other Ruby code, such as my concept code,) and can only do so when a certain scene is the active scene page.
This code, and the Dynamic Components Ruby code is not running when you are viewing your viewports (of model scenes) inside LayOut. So the DC door would appear as it does when statically saved to the SKP file. (Which is a closed door.)
So, the only solution is, inside SketchUp, to make the door or drawer a component. Have two instances on two layers, and two scenes. One for closed one for open. Copy the closed instance in place, and then transform it as needed (slid outward, or rotated on it’s hinges, etc.)
I want to thank you for bursting my bubble. Seriously, now I know that what I want isn’t possible, and I can turn my attention to something that will, and not waste any more time on something that won’t.