Yes, I can test this, but I was thinking along the lines of "designed in" extra CPU cycles that may be in the TCP connection verses the local protocol connection. If I'm going Local database to Local backup, I'm hoping that there is no difference or if there is, that it's negligible.
Depending on the theoretical difference, if it's small, it wd take a very large database backup to appear as a significant difference in performance.
Put more simply: Is TCP protocol slower, or have more CPU overhead, than Local protocol, if both the database and the backup file are on the same local computer?