Hypothetical continuous query answering over data streams was introduced as a way to anticipate answers to queries that depend on facts that may or may not happen in the future. Previous work has studied this problem for Temporal Datalog with negation and instantaneous communication, showing that hypothetical answers can be incrementally updated as new data arrives at the data stream. In practice, individual communications take variable amounts of time, so data may arrive delayed and unordered. This motivates studying hypothetical continuous query answering in a setting with communication delays. The interaction between communication delays and negation is however problematic, and the existing approach is restricted to the positive fragment of the language. In this work we show how to remove this restriction by defining an appropriate operational semantics based on fixpoint theory, and showing that the relevant fixpoints can be computed in finite time by a carefully designed algorithm.