Implementing Data Virtualization for Unified Data Access

Assume the role of a data integration consultant and provide a detailed strategy for implementing data virtualization to achieve unified data access across multiple databases. Include the following elements: guidelines for selecting a data virtualization tool, steps to integrate data from sources like PostgreSQL, MySQL, and NoSQL, techniques to ensure data consistency and security, and best practices for optimizing query performance. Outline potential challenges and solutions to ensure efficient access without data replication. Target the strategy for a medium-sized enterprise with existing disparate data systems, focusing on scalability and ease of maintenance.