Support was able to help. The problem was simple, identifying it not so much. Even though the snapmirror relationship had been removed entirely, it still wasn't broken. I'm not sure how that works, nor does it matter. Simply running snapmirror break allowed me to then delete the root volume:
mycluster::> snapmirror break -destination-path mydrtest:
mycluster::> volume delete -vserver mydrtest -volume mydrtest_root
Warning: Are you sure you want to delete volume "mydrtest_root" in Vserver "mydrtest" ?
{y|n}: y
[Job 1806] Job succeeded: Successful
