There is always room for differentiation at the top, either as an architect, or leader - but (speaking of my own profession) - I've never actually seen an environment where any particularly network engineer wasn't replaceable. In fact, one of the hallmarks of a first class network engineer is that they leave behind an environment that is well documented, designed to industry standards, and is pretty much resistant outages as a results of single points of failure.

A Network engineer who is a 'Hero' and can never go on vacation, is worth much less than a network engineer who has built an environment that any random Taos CCIE can come in and manage with a couple days cross training.

Ironically, from that perspective, the more fungible [1] a network engineer is, the more valuable they are.

[1] http://en.wikipedia.org/wiki/Fungibility

You seem to be contradicting his point, but as far as I can tell you're proving it.

The fact that you see such a stark difference between a first-class network engineer and the work of lesser mortals is proof that they aren't fungible.

I'll also point out that you have, perhaps unintentionally, changed the topic from design to maintenance. We're talking about product creation. Some people like to pretend that all the design happens in the head of the person with "designer" on their business cards. But anybody who has spent time in a team making great products knows that everybody is involved in design. It can't be otherwise, because design happens at the intersection of desire and possibility, and the technologists know the possibility space more deeply than anybody else can.

That a well-designed network can be easily maintained is demonstrates the great design skills of that first-class network engineer. That there are long periods in a network's life when design skills aren't needed isn't proof that technologists are interchangeable; it's proof that they aren't.

I'm not sure why you don't apply that reasoning to managers. You do not need continuous guidance from the CEO.

A great CEO, after he has built a good team and a good strategy can just let the company sailing along for a decade. He is just as fungible as the network engineer. I mean, Jobs died and Apple didn't immediately go bankrupt, another CEO just stepped in, and either cruise along ( or picked up where Jobs left of, future will tell )

The big difference between the 2 is the immediate impact and that's more a cultural thing than anything. If you look at it from an engineering point of view, you avoid at all cost to design anything with such a fragile, expensive and exposed element as is the CEO role in a business the size of Apple.

I left the field because it is a pretty inglorious role: nobody congratulates you when the network withstand the load under the worst circumstances, but everybody point you a finger for the minimal glitch - when it is not your fault you have to prove it.