[abenedik]

I was thinking that you would define 2 ModelVisual3D objects - each would have wireframe and transparent mesh. This would require using VisualEventSource3D instead of MultiModelEventSource3D (I would create two VisualEventSource3D - one for each ModelVisual3D).

So when you would change selected object, you would register the VisualEventSource3D of the selected objects and add other Visuals to exclusion list with RegisterExcludedVisual3D (note that you can use RemoveExcludedVisual3D method to remove the Visual3D from exclusion list).


Another option:
What if you would only add one transparentMesh to the Viewport3D - one the one that you would like to be detected. The other transparentMeshes would be removed from Viewport3D?