Sounds partially correct, although, I am a self-taught "economist" mostly.

However, in my experience, this kind of "basic incentive structure analysis" type of economics does not really predict the real world so well. I had to actually un-learn a lot of it in real life. (Or maybe I should have learned more advanced kinds of economics.)

Basic Econ can be understood as more normative than descriptive. That is, if you had these kinds of markets, then things would be really great and efficient. But in real life everything from old-boys networks to political corruption to cognitive biases screwing this up. The more you live in a place that works like Basic Econ the closer you are to an optimum. If you see lots of poverty around you, someone, something is throwing big monkey wrenches into this mechanism.

So in the real life you can see well-paid employees who are just nephews of the CEO or have an excellent sales pitch (interview skills). It does not always work out the textbook way. It should, just doesn't.

Your job is probably a commodity. Engineer, baker, tailor, data scientist, statistician...it's a commodity unless you can make buyers believe that other sellers are not selling the same thing that you are - basically, that they can't approach other sellers and get a price from them for the same good.

But most people are specialists today. When was the last time you saw a General IT Guy in a big city? 1990? Today it is being some kind of an Enterprise Network Bullshit Bingo Database Architect. Or more realistically, simply based on what kinds of products or technologies they know. For example being an Oracle Financials expert is a world on its own. Or SAP, Navision...

But if you become a specialist, there is one trade-off. You must live in a city. Rural areas need generalists. This is simply based on density. In the country, there may not be more than 5 doctors or IT folks within 100km so they better know it all.

I think that large companies likely have quite a bit of inertia, meaning that they may not be making "efficient" decisions either. The people working at those companies have TVs that are probably risk-averse with respect to money. Firm capital (if publicly traded) depends on shareholders transient judgment...which may be to SELL due to their own risk aversion if the firm makes a radical move. So I would expect firms to be not efficient in the strict sense, and also risk averse. So it looks like there is a tradeoff as firms get bigger (as revenue increases and costs increase). They likely get some smoother processes, less hiccups, economies of scale etc, but they also possibly give away expected utility through risk aversion.

The big issue is not even risk aversion. After all a truly large firm can risk a tiny fraction of its revenue on an experiment and that tiny fraction can still be large. The issue is plain simply the incentive chains. There is a long hierarchy with everybody on higher than you trying to use rewards and punishments to get you work the way they want you to. After several steps down in this chain, this naturally gets corrupted. Like a game of telephone. People who are 4-5 level removed from shareholders they don't want the same things shareholders want to. If they are rewarded for number of calls taken, they will find any excuse to drop calls and take another one.