What makes software developers "so much more special than " is that their output to the business is teaching a dumb-but-very-fast-and-cheap-and-capable machine how to do a previously-only-done-by-humans task. It turns out that's a
very valuable thing for businesses to have. All those other trades are in "exploit" mode on the "Explore -> Expand -> Exploit" spectrum. There was a time when having a skilled carpenter was critical to building your company. But that time has largely passed. Most carpenters today are applying known solutions to known problems.
Another way of saying the same thing is that few businesses can see their revenue double by employing even the most skilled plumber. However, many businesses can see there revenue double (or their costs halved) by employing even a mediocre software developer.
So, I would agree it's true that we live in a special time in history where the translation of business needs to computer languages is being demanded at an ever higher and higher rate. But I would disagree with the framing of this phenomenon as a "tech bubble" or a "distortion on commandable salary".
Employing a software developer IS expensive.
Another way to look at the status quo is to see that the beauty of capitalism is that those areas that consumers see to be the most important areas to focus on having software developed ARE having software developed. Software developers are working on the most important projects they can be working on
right now. However, the problems software developers are working on will change over time as:
- Existing problems get solved.
- New, more important, problems emerge.
- More developers enter the market (bringing down the
cost of development).
- Consumer demands shift over time.
On that last point, I think it's interesting that the behemoths of software development are all seeing huge tectonic shifts in their industries occur. FB/Meta responding to Apple's/EU's/California's privacy enhancements. Google desperately trying to find a non-ads source of revenue, knowing that their cash cow could dry up in the not-too-distant future. Twitter being on the bleeding edge of discovery of "How much content moderation is too much?". Netflix figuring out it's tech moat wasn't as wide as it thought it was. Amazon...well, I'm not sure what tectonic shift is hitting Amazon yet, but it can't be far off.
How many of these companies will be around in 10 years in the same form that they are today? Ford has been making vehicles for over 100 years. What will give Meta, as an arbitrary example, that kind of staying power?
As those companies come and go over the next 10-20 years, I think your concern that "the rest of the economy could really use software developers" will be ameliorated. Maybe everyone's actually working on unimportant problems right now, and the wider economy has more important problems for us to solve, but the best and fastest way we've figured out how to discover that we've misallocated resources (whether human or any other kind) is to let everyone make their own decisions and realize the individual gain or loss of their decisions.