Failing that, backgrounding the process at launch (by appending & to the invocation) should stop it blocking the shell from which it's launched.
While I did mean what I said earlier about not feeling qualified to argue with someone who's been using Emacs since before I was even a gleam in my father's eye, I do want to point out that it's maybe sort of unfair to blame this one on Emacs; any process you launch from a shell, which doesn't immediately perform a detaching fork from its parent process (as most things don't), is going to have the same effect, and require the same modified invocation to prevent blocking your shell.
Does emacsclient -n not work?