Which will do nothing, because it doesn’t actually tell you if the peer is connected - it’s literally just telling you the name of the signal. That’s not how you use signals. You cannot assign them to variables and do anything with them. (Other than potentially obfuscate your code.)
You’re not doubling up on signals.
If you wanted to get something useful out of that signal you’d do something like this:
Having said all that above, the answer to your initial question is: No. The interpreter is going to optimize your code better than you can. So whether you obfuscate your signals, or call one function with a reference it could have pulled in code (as @normalized pointed out)- it makes no difference. It’s going to run at the same speed. The only thing you’re doing in your code examples is make your code much harder for you (and anyone else working on your project) to read.