The "exhausted low hanging fruit" model IMO has been wrong since the industrial revolution. The fundamental problem is it's based on the heuristic of fixed demand + technology saturating the demand. It's really the opposite - technology is an ever expanding fractal, and the larger the surface area, the more things are needed. For example compilers are having a huge boom now because hardware is increasingly performance demanding and heterogeneous.
It's easy to find new uses for technology. The hard part is turning the new idea into a commercially viable product. And it's even harder to find a business model that doesn't destroy the product in the long term.
The whole is often more than the sum of its parts. A company is not just a bunch of people, but it's also the company culture and the established ways of doing things. It's common that a company can't make something work, but the same people in another company could, because they can organize their labor in a different way.
Software development is often an investment. Hiring a developer is often an investment. Most companies eventually reach the point, where the rate of investment slows down and the company chooses to return the profits to the shareholders. When the company no longer believes that it can invest the money more profitably than the shareholders could, it's rational to lay off people working in R&D and use the profits for dividends or buybacks.
Dividend buybacks aka stock manipulation. You are saying that it is logic that companies turn from making things to stock manipulation. That is a system that just doesn't work long term. I guess you are saying companies have a life cycle, and once it reaches the stock buyback phase they are basically dieing.
