polyglot-kotlin docs still mention KT-27956

[seanf]

The docs for polyglot-kotlin still mention the Kotlin issue KT-27956, but this issue and the follow-up KT-33892 are marked as fixed since Kotlin 1.3.60.

Is the performance impact still present?

[seanf]

@lion7 @thorntonrp
Am I right in thinking that your changes in #184 probably took care of the performance problem mentioned in the readme for polyglot-kotlin? That PR seems to have eliminated all references to ScriptEngine.

polyglot-maven/polyglot-kotlin/README.md

Lines 96 to 112 in 5a92df1

	## Known Issues
	- Each Kotlin `ScriptEngine` requires a classpath that includes information from Maven's own classpath as well as the
	extension's `ClassRealm` (which is a special type of `ClassLoader`). This classpath is used to create an additional
	`ClassLoader` for use by the `ScriptEngine` which requires the Maven process to consume more metaspace memory than it
	would if the Kotlin `ScriptEngine` were able to obtain the class information it needs from the existing classloader.
	This is a limitation of the Kotlin script engine. (See [KT-27956](https://youtrack.jetbrains.com/issue/KT-27956))
	While this issue does not appear to have a significant impact on Maven's command-line performance, it does affect the
	performance of IntelliJ IDEA's remote Maven server process (as of Build #IU_183.5912.21). The IDE's remote Maven
	server appears to create a new `ClassRealm` every time it reimports the Maven model. This in turn causes a new
	`ScriptEngine` with its additional `ClassLoader` to be re-created even though these are created as singletons. This
	results in the IDE's Maven server process consuming much more metaspace over time than it would otherwise, especially
	for multi-module projects.
	If you care about this issue, please vote for [KT-27956](https://youtrack.jetbrains.com/issue/KT-27956) under the
	Kotlin project so that the Kotlin script engine can get the information it needs from the existing classloader.

[lion7]

I think you are correct, the performance problem should be gone now that the ScriptEngine is no longer used.
Maybe you could create a PR to remove this known issue? The one thing I'm not sure of, is how to measure that it is indeed gone. I'm hoping 1 of the maintainers has a more solid opinion on this, @mosabua perhaps?