We actually discovered this a few months ago, thinking it was an issue with our SSO setup for our Read-Only users. After talking to our reps about it, they confirmed that the troubleshooting tools are available by design for Read-only users. We submitted feedback to change that, as well as at least log a port cycle since that's kind of important.
What is extra annoying is we have the ability to 'tag' ports with a description and tie configuration changes to specifically those ports via SSO / Azure roles, but it doesn't stop the port cycling of other ports as soon as they have Read-Only rights to the network.