You would need a target host outside the firewall that would react in predictable ways to determine if your outbound policy is working correctly. It would be hard to determine if a random ephemeral port, like 46622 is responding as closed because the firewall blocked the connection, or if the port was actually closed on the remote host.
For something like this, I would probably resort to netcat and do some scripting. Or maybe a tool like Firewalk, which can test lots of permutations but is EXTREMELY slow.
Normally outbound rulesets are so sparse that a manual approch is not unreasonable.
pansophic