@HeavenlyPossum
This is the key bit. The theory is that people who are good at allocating resources will end up with more resources and so will allocate more resources more efficiently.
There are a couple of problems with this. The first is nicely summed up by a game where you all start with $100 and are allowed to bet up to 10% of your wealth on the outcome of coin flips. After the first round, some people will have $110, some will have $90. People who won in the first round but lost in the second will have $100, but people who lost in the first round and won in the second will have $99. People who win a few early rounds are able to accumulate a lot of the wealth in the game. They are not good at predicting the outcomes, it’s entirely random, but eventually this game decays to a distribution where a few people have most of the money and most other people have none.
The second, as you point out, is that extreme concentrations of wealth insulate you from failure. A trillion dollar company with multiple revenue streams can fund all sorts of things that waste billions of dollars, skew other markets, have huge externalities (environmental or social damage, for example) and have basically no consequences. Meta is a $1.5T company. Their stock price is around double what it was before the start of the pandemic. Wasting $70B didn’t make them lose money, it made them make money less slowly.