The behavior you are describing is odd, so we are looking into it to see if we can reproduce what you are experiencing.
The IP should not go into a Transient state if if responds to ping, but not SNMP. It should only go into a Transient state when it was responding to one of those methods and is no longer responding indicating the IP may now potentially be open. By default we set Transient state at 7 days (which you can change), after which if we get no response for that time period, we return that IP backed to Available. Hopefully that explanation makes sense?
I will let you know what we find on our side and if a defect, one will be filed.
Ok, let me know if you want me to do anything, I was thinking of removing the SNMP communities and scanning again, but haven't so far.
Or if there is any logging we can turn on to track the logic somehow, let me know... I'd really like this ability in an end-product BTW. The ability to turn on debugging and log things is always a good thing!!