It's putting a packet through a 150w CPU like weighing a brick is using using a 500 lb scale. I.e. just because it can do that much and you're using that device doesn't mean it's the relevant number to look at. When you're not pushing things it'll be near idle wattage, when you are pushing things like mad it'll be around <4 core wattage (which will still be a lot less than 150W) in this workload. The important thing is it's a CPU which has a great single thread performance and, as a result, it has a high peak multicore wattage you won't actually care about.
Overall I'd bet all of the NICs and transceivers are the majority of the power usage in a given year.
I still need to measure and optimise power consumption correctly, but it drew about 70W from the socket during the tests, according to my simple power meter. CPU never ran at boost frequency, and a lot of other stuff is in a low power state.
Overall I'd bet all of the NICs and transceivers are the majority of the power usage in a given year.