I have noticed that defining identical property names in sibling classes seems to be causing problems when you instantiate any of the two sibling classes. 

Example: 

class A <- super class

class A.1 <- inherits A

int troubleCounter <- property in A.1

class A.2 <- inherits A

int troubleCounter <- property in A.2

This defintion would pass validation and rebuild, but when you add an instance of either A.1 or A.2 you will get an error message stating something along "key was already added".  Does this mean that properties in A.1 and A.2 are not isolated from each other?  Wouldn't it be useful to be able to declare similar properties in a subset of an inheritance tree, without having to add another class layer to propagate the property to the wanted classes?