First, ALL of the below is based on the presumption that the zone has been properly delegated.
After several tests, retests, pondering, rinsing and repeating, here's what I'm observing:
If I do a generic query for a PTR record of one of the above four addresses, I get mixed results, depending on which of your DNS servers was allegedly queried.
Sometimes, but very rarely, I get a positive response from a generic query.
If I force the query to a specific DNS server (I tested both 158.106.84.10 and 158.106.84.11), I get valid responses every time.
This is the result of a forced query to 158.106.84.10. Only ONE response is an expected condition in a directed query.
But here, I sent a generic query. It failed.
FWIW, I do not have this behavioral issue when querying my own IP Address, the IP Address of DNSStuff.com, or the IP Address of Solarwinds.com, so I do not believe it is an issue with DNSStuff.com
I would suggest monitoring the traffic across port 53 while initiating another series of queries, and see if ALL of the queries are crossing your threshold.
Also, enable logging on all of the DNS Servers and see who/if/when queries are received and/or responses are generated.
The missing link at this point is identifying exactly when/why the lookup fails.
