RESTful web services and APIs are popular in industry and represent one commonly used way to expose functionality via a well-defined and technology-agnostic interface. While these APIs can be analyzed based on best practices or antipatterns, their interface quality can also be evaluated with metrics. Several of the interface metrics proposed in literature have been implemented in the RAMA approach based on API documentation. To empirically evaluate the effectiveness of the proposed interface metrics, we analyzed a large sample of publicly available APIs and compared the metric values to feasible ground truths for software quality. API descriptions were analyzed using the RAMA CLI and software quality metrics were collected using SonarQube and git. We used multiple linear regression models to examine the correlation between API metrics and software quality metrics. Furthermore, we studied the trend of the correlation over the evolution of a project. Our results suggest that some API metrics statistically significantly correlate with maintainability metrics. However, the regression models and the trend of the correlation indicate that as a project evolves, an increasing number of factors besides the API metrics influence the quality of the source code.