I didn't read the journal article, but based on this post, science didn't explain WHY anything. (i.e. why the silent treatment works... and why it's healthy). In this case, as in many cases of experimental work, science shows that a correlation exists. But it didn't describe why the correlation exists.

It's because limiting the number of logins or the rate of logins doesn't prevent other password attacks where the thief can obtain your password without trying to login.

Maybe it's obvious, but your situation is exactly why hackers go to great lengths to hack a single password: If any single site that handles your money leaks your password, then ALL the accounts - PayPal, your email, your online banking - that have the same password are immediately compromised! Those sites might not…

So, if someone hacks your passphrase, GAME OVER for you.

Yes! I use this kind of thing all the time: For example, check for core dumps in a bunch of remote servers:

This. GNU screen. Multiple terminals inside one terminal. If your Internet connection breaks, your terminals are maintained; you can reconnect and your terminals sessions are just the way you left them, still running. You can split the screen to show multiple terminals. You can have screen alert you when there is…

Probably the best assurance that there's no hanky-panky stuff going on is Open Source.