After SERVER-119295, ExplainCmd and ClusterExplainCmd will be typed but it is still not leveraging the full benefits of TypedCommand and IDLs due to having a raw BSONObj representing the "inner" command that is being explained and thus requires some manual parsing for fields such as "$db" and "comment". We should investigate whether we can model all explain-able commands as a variant in explain.idl so that we can get stronger types for the various explain-able commands.