I think there's probably a lot of precedent for "seasoned replacements" making a positive impact when it comes to technical infrastructure. Not that I can point you to any. But my sense is that most big sites on the web owe a good chunk of their scalability to some experienced ops genius(es).
That could be the case at Facebook. I had the opportunity to talk with Jeff Rothschild, Facebook's VP of Technology, last year. He was very obviously both experienced and smart.
