The most significant downside of using calculated dimensions in a large, data-heavy workspace is performance. Calculated dimensions are computed "at query time," meaning the calculation is performed every time a widget, dashboard, or report that uses it is loaded. In a workspace with one hundred data streams and a large volume of data, these on-the-fly calculations can be computationally expensive. This leads to slow query execution, increased dashboard loading times, and a degraded user experience, which is a more immediate and pervasive issue than setup, maintenance, or the abstract concept of scalability.

B. Ease of setup: Setting up a calculated dimension is a relatively straightforward process within the Marketing Cloud Intelligence user interface and is not considered its biggest drawback.

C. Ease of maintenance: While maintaining numerous calculated dimensions can become complex, the constant, system-wide impact of poor performance on all users is a far greater downside.

D. Scalability: Poor scalability is a result of the performance issue. The direct problem is the performance degradation itself, which becomes more apparent as data volume scales.

1. Salesforce Help, Marketing Cloud Intelligence Documentation, Calculated Dimensions and Measurements in Marketing Cloud Intelligence. This document states, "Calculated dimensions and measurements are calculated during query time." This confirms the technical reason for the performance impact, as the logic is not pre-processed during data ingestion but is executed with every user request.

2. Salesforce Help, Marketing Cloud Intelligence Documentation, Best Practices for Query Performance in Marketing Cloud Intelligence. This guide explicitly advises on performance optimization. It recommends minimizing the use of complex, query-time calculations (like calculated dimensions) in favor of processing logic at the data model level (e.g., in the data stream or using Data Fusion) to improve dashboard and report loading times, especially in workspaces with large data volumes.

3. Salesforce Help, Marketing Cloud Intelligence Documentation, Create a Calculated Dimension in Marketing Cloud Intelligence. The "Considerations" section of this guide often highlights that while powerful, these dimensions can affect performance. It recommends testing their impact in a staging environment before deploying them widely in a production workspace with significant data.