I think there are valid use-cases for having an object in multiple groups. E.g. you have two groups that you alternatively hide. The object in both groups will stay visible. Think of a multiple button-bars, all with a shared OK-button.

In your example you are adding the object to the same group twice. That is a bit different and might be considered as an error.

For the re-used 'Okay' button I think a Bitmap and TileGrid are more likely to be in use than a Vectorio shape. In the case of Bitmap and TileGrid you can have multiple TileGrids that share the same Bitmap, so you could make two TileGrids that re-use the same Ok button graphic Bitmap file and put them in two different Groups.

I haven't traced through all of the code to be sure, but I think that both TileGrid and Vectorio shapes have internal variables related to dirty region and refreshing that are used by logic that is written to assume that there is only a single instance of the thing, and it will be in only a single Group.

If a thing were in multiple Groups it would need to keep straight which instances of itself are dirty and need to be re-drawn on the display whereas I think right now it has only a concept of dirty or not, and not have a concept of "is this instance dirty, and is that other instance dirty independently."