I prefer to keep the code as-is.  It is microscopically faster than than just inheriting the methods, but more importantly it makes it easier to read the code and know that it is correct.  We're instructing people to override all four methods when subclassing and we don't make any promises about how the parent method is implemented.  At this point, the implementations happen to be the same, but that isn't guaranteed.  We could at some point change the implementation of the parent class method and wouldn't want the subclasses to break.

So, let's keep the loose coupling, the self-contained code that follows the rules, code that is slightly faster, and code that is easy to read.

Thank you for the suggestion, but this micro-factoring doesn't make us better off.