What would be the advantage of using
can.Observe.setter
instead of
"obsrv.bind( 'change', function( e, attr, how, newVal, oldVal ) { . . . "
and writing change functions for needed attr targets?