I can see how it is similiar to PUP-10118. Is the resource purge logic independent from the resource provider? I can't think of how a cron resource wouldn't be idempotent.
Another thing to keep in mind, from what I recall the cron purge logic also isn't working correct for when a cron entry exists in the wrong cron user tab. I feel like the user value should be part of the namevar.