Value Classes Still Need Compiler Sympathy

JDK 28's preview of value classes (JEP 401) offers significant optimization opportunities by allowing the JVM to choose appropriate memory representations without maintaining object identity, but developers must understand that value classes don't guarantee performance improvements and can actually underperform ordinary classes in certain scenarios when the JVM must convert between flattened and reference representations. The key optimization advantage depends on immutability and strict field initialization, which allow the JVM to safely flatten values and avoid pointer chasing, but mutable fields or situations requiring materialization can force less efficient reference layouts.