I think you can see some of the struggles even in programming. I've spoken with managers who are honest about wanting replaceable coders, and won't willingly choose a technology which swings the balance of power away from them. (Which means they'll likely prefer Java over certain other contenders, even if they actually happen to believe this replaceable-cog strategy costs them more time and money, thereby lowering profits.)
(That's why many are attracted to startups; some of these counterproductive constraints are relaxed, due to the sheer difficulty of success.)
I think all things equal, you do want to be able to be worker independent: you'd like to have everything written in, say, Python not a mix of languages each developer happens to like. This doesn't mean that devs are actually plug-n-play, but on a spectrum is does make them more so.
From a pragmatic point of view, I'm sure subjected to blatant commodification, a lot of people will react by quitting or creating systems that are hard for new people to pick-up: i.e. obscure / spaghetti code.
In reality, most engineering jobs are pretty far from commodity - especially if you'd like to keep your team cohesion, tacit knowledge, experience, etc. It generally doesn't pay to do this in creative jobs.
There were a lot of points / quotes in the book with management talking about reducing / commodifying labor, but I'm not sure this book is the best to cite as support for this. I suspect Chomsky is taking Nobel at his word, but the book is noticeably biased toward Marxist thinking in thinking everything is a class struggle whereas the technology actually was superior.
Funny enough the operator-programmer lines are blurring today as CNC code is now generated on essentially a compiler/debugger that simulates the machine. Machine programmers are scarce and now (even sans unions) are actually able to demand high prices in the same way a web developer can in our business.