The point of using an optional is to signal that the value may be ``++nil++`` and to provide graceful ways of dealing with it if it is ``++nil++``. While implicitly unwrapped optionals still provide means of dealing with ``++nil++`` values, they also signal that the value won't be ``++nil++``, and unwrap it automatically. In addition to sending a decidedly mixed signal, this could lead to runtime errors if the value ever is ``++nil++``.
It is safest, and clearest to use either an optional or a plain type and avoid the boggy middle ground of implicitly unwrapped optionals.