Note: We maybe should use the --plugin option with care. It will also prevent the core javascript tests from running. But imho it can be useful to know if the core javascript tests are still running if a certain plugin is used.

Why would core javascript tests need to be run in a plugin build? Or why would we need to know if the core tests are still run in the build for a plugin?

A plugin might break the core tracking even with having succeeding JavaScript plugin tests. Running core tests might show that.

How would it break core tracking? Through tracker events? Or something else?

Maybe. That was more a random thought, than something that would have happened in the past.
Being able to be sure, that tracking tests of core succeed with a plugin, is imho better than not knowing it.

Failure tests could be added to core in that case, ones that simulate broken plugin behavior and make sure the tracker survives? And if it's really an issue that plugin JS can somehow break core tracker JS, core tests can be grouped together (if possible to do a module within a module in qunit) into a core one which is simply run with another tests:run-js command before the --plugin=PluginName one. Though it's hard to see what could actually go wrong given core tracker code should not interact with plugin tracker code directly.