Well part of the benefit is rapid development; it's mind-boggling how quickly someone can stand up a dbt project and begin to iterate on transforms. Using Python/SQL/JSON (at small/medium) scales keeps the data stack consistent and lowers the barrier to entry. No reason to prematurely optimize when your bottleneck is the modeling and not the actual data volume.
dbt and ELT in general are such a game-changer for allowing rapid iteration on business logic by data analysts; the feedback loop feels much more like "normal" software engineering compared to legacy systems.
Still not sure whether this is serious or not, but it's not really infrastructure as SQL, it's infrastructure as database records which is stateful and defeats the point.
The rate of return on investments is proportional to the risk and/or effort necessary to make it work. Contracting out the management of things like condos, vending machines, etc. destroy the vast majority of any gross profit. Investing in a large-cap, high-dividend mutual fund (Vanguard, etc.) will generate reasonable returns with 0 effort on your part. Just cash out your dividends rather than re-invest.