Designating all elements of the DMEL as optional may create a problem; a minimum set should be required.
The user may need some extrinsic characteristics of the resource such as its cost and hence should be included (though optional). All documents of any permanence should be archived and accessible.
Subject element should be made up of two sub-fields, a schema and an hierarchical subject field which includes sub-subject and sub-sub-subject. The classification scheme used (authority) could be specified by the schema field entry specifying the cataloging schema. The hierarchical subject field entries must be from the same schema.
For non-titled objects, we need an algorithm to insert a(n alternate) title. For example the non-titled resources such as satellite data can have a cooked up title using satellite name, time, date, position, camera orientation, and filter or frequencies etc.
Relation and source have similar semantic implications and could be described using a relationship subfield with the identifier of the object to which it is related. An optional sub-field may be used to provide annotation or useful information.
In addition to the language of the DLO, the (natural) language used to specify the elements of the metadata and the character set used have to be specified. Furthermore, elements such as an abstract and annotation are missing. The former, generated by a human or other agent gives a capsule idea of the contents of the DLO. The latter is a place-holder for additional details regarding the DLO by responsible agents or other users.