Guard interfacial water entropy for solute-free systems and fix WaterEntropy reporter usage

[harryswift01]

Summary

This PR fixes a crash in water entropy calculations for pure solvent systems and improves correctness of water group detection when atom selections are applied.

Previously:

• Water entropy was computed even when no solute existed, causing the interfacial solver to fail.
• Water detection used the full universe instead of the reduced selection universe.
• WaterEntropy was instantiated incorrectly without a reporter.
• The wrong method was invoked when computing water entropy.
• This PR resolves all of the above and improves workflow robustness.

---

Changes

Interfacial Water Entropy Guard

• Water entropy is now computed only when both water and non-water groups exist
• Prevents interfacial entropy from running on pure solvent systems
• Avoids MDAnalysis selection failures caused by empty solute detection

Universe Consistency in Water Group Detection

• _split_water_groups() now operates on the same universe used to generate molecule groups
• Prevents mismatches between grouping and water classification
• Ensures correct behavior when reduced atom selections are applied

Reporter Integration Fix

• WaterEntropy is now instantiated with the required reporter:
• WaterEntropy(self._args, self._reporter)
• Fixes missing dependency required for logging results

---

Impact

• Fixes runtime crash in pure solvent simulations
• Improves correctness when using atom selections
• Makes water entropy logic consistent with interfacial definition
• No impact on mixed solute-solvent systems

[jimboid]

This PR fixes an issue caused by renamed functions. It also guards the usage of waterentropy for interfacial water calculations when there are no water based systems in the molecular system.