Is this a tcpwrapper bug?

[Tom Horsley]

While trying to make my ssh connection maximally secure,
I decided to use the /etc/hosts.allow and /etc/hosts.deny
files, and was able to get them working without too much
trouble, so now the only external site that can connect to
my home system is the address of what is probably my work
system's firewall.

Here's the bug: The DNS setup for the IP address of the
firewall gives it two names, both (examples only)
users.example.com and corpvpn.example.com are in the
DNS database as the same IP.

If I try to use either name (or both names) in the hosts.allow
file, I always get rejected with a log message that claims
users.example.com != corpvpn.example.com.

If I use the IP address in hosts.allow, I still get the message
in the log, but I am allowed to connect.

Is the tcpwrapper code being too paranoid here? Shouldn't
it lookup all the names for the IP and accept any match
rather than (appapently) only looking up the 1st name?