What is your threat model that requires such complexity?
It seems like privacy cosplay to me, tbh. I'm genuinely curious what type of threat it actually protects against, if any. What happens when two different systems you ssh into knows it's elric on both?