@nbrinks 

NAT Boost is supposed to redirect the CPU to use processing power to make speeds faster, this applies to wired and wireless. If you are getting 200 Mbps, I would not consider 160 being slowed. If you more then half I would begin to get worried.

You can temporarily change the wireless password or turn off Wi-Fi, and only have the computer wired to see if the other devices contribute to the drops. I would also use other wired devices to test the speeds to see if the drop just happens on the computer.

@Tony 

Thanks for the reply.

I have changed my conclusion about wired/wireless. After some additional testing, I am now uncertain if wired/wireless is a factor.

I am still seeing a difference between NAT Boost On/Off however. Strangely enough the difference in bandwidth for NAT Boost On/Off appears to be highly dependent on the server I am testing against. After more than 30 tests, using two different computers (all gigabit ethernet), against multiple endpoints I saw a range of bandwidth drops between ~0% and ~70%. The difference in bandwidth is most dramatic when performing single connection bandwidth tests. 

I don't have a lot of time to dig into this any more but I am consistently experiencing the highest downstream bandwidth with NAT Boost Off (disabled), at least according to several servers via Speedtest.net.

I am just going to keep NAT Boost off. I only bring this up because I thought NAT Boost (enabled by default) was supposed to improve performance, however, my experience suggests the opposite. Perhaps some layer in my network stack doesn't behave well with NAT Boost but these are the results I am seeing at least.

Thanks!