The critique of the Apache Iceberg REST Catalog highlights significant flaws in its operational reliability, despite achieving semantic clarity. The specification provides clear definitions for metadata operations like tables and namespaces, allowing different engines like Trino, Spark, and Flink to communicate effectively. However, it fails to address critical performance aspects, such as latency and throughput. Without operational service-level agreements (SLAs), systems that comply with the spec can vary drastically in performance, leading to unpredictable behavior in production environments. One major issue arises with the ListTables endpoint when dealing with multiple catalogs. For example, if an Iceberg table is registered in both Catalog A and Catalog B, the synchronization process can become burdensome. As tables multiply, sync operations can take minutes or even fail outright, causing delays that disrupt query engines. The absence of defined expectations for how long these operations should take forces clients to adopt defensive programming strategies, inflating load and exacerbating latency issues. The article also examines the implications of optimistic concurrency control in Iceberg, where simultaneous commit attempts can lead to conflicts. The specification only mentions a conflict response but lacks guidelines for handling retries fairly, which can lead to issues like starvation for certain jobs. Furthermore, caching is poorly defined, leaving clients to frequently re-fetch data instead of relying on cached responses. Overall, the lack of operational guarantees results in a system that, while semantically correct, is fraught with reliability challenges, particularly as Iceberg transitions from a testing phase to core infrastructure.