The debugName field is useful for debugging java.lang.invoke, but is generated also for production code and has been used to help name the actual invokers. This is adding some footprint and startup overhead, while the naming of methods and classes generated could use the Kind of the LF.

There's also a subtle issue in that when editing a LF the debugName is passed along to the new LF, which obfuscates which LF is actually what.

Removing debugName in favor of a debug name association that's only generated when actually debugging should be considered.