Ensuring the provision of consistent and irredundant data sets remains essential to minimise bugs, promote maintainable application code and obtain dependable results in data analytics. A major challenge in achieving consistency is handling incomplete data, \ie missing information that may be provided later, comes from the fact that, the use of marked (or linked) nulls is required in many applications to express unknown but connected information. In this context, it is well known that maintaining the data consistent and irredundant is not an easy task. This paper proposes a query-driven incremental maintenance approach for consistent and irredundant incomplete databases. Can graph databases improve the efficiency of this operation? How can graph databases manipulate linked nulls? What is the impact of using graph databases on other essential maintenance operations? This paper presents an innovative approach to answering these questions, highlighting the proposal's strengths and weaknesses and offering avenues for further research.