That is what I would expect. I see throughput as what is actually consumed by the client and the goodput as what the client is expected to reach if it would be sending and receiving data all the time (like the reasonable expected peak throughput as goodput and actual data used as throughput).
I found this explanation op goodput a nice one. I see goodput as the link speed minus the overhead minus inefficienties like retries and such, resulting in a value that can be actually achieved.