After some thought I ended up just failing the delete with a more clear message as a reason of failure. AFAIU we can not warn the heatclient user without failing a delete, and leaving the container to exist after successfully deleting the stack will leave a presumably billable resource behind, which we probably do not want (as for TempUrls they have limited time span, so while still not ideal, the user would not end up with unlimited sum billed for them).
After some thought I ended up just failing the delete with a more clear message as a reason of failure. AFAIU we can not warn the heatclient user without failing a delete, and leaving the container to exist after successfully deleting the stack will leave a presumably billable resource behind, which we probably do not want (as for TempUrls they have limited time span, so while still not ideal, the user would not end up with unlimited sum billed for them).