> I'm excluding individuals getting rich as that can happen due to gambling, etc.
The tools that allow that stuff to occur onchain are pretty phenomenal. Bridges, farms, yield aggregators, yield optimizers, AMMs, the rapid evolution of standardized classes in the codebases (really was a mess before OpenZeppelin). It is capital formation and that is valuable for many people, a lot of capital formation improvements in tradfi never caught on for failing to address more important pain points people encountered.
For the aforementioned tooling, many of these services get hacked and take a lot of the capital inside of them, and many of them function really well and still are to this day. The velocity of permissionless deployment lets the market/conscious quickly see which ideas work well, and become a building block to try something else with. (I say "conscious" because the developers often aren't the market).
The best example of a recent reporting asymmetry would be all the stablecoins that aren't failing, functioning seamlessly, and what can be done to improve even them for a model stablecoin. Reporting about active proposals and the progress being done. This stuff is reported, just not in mainstream news sources and it would be an okay fit as some of them also have dedicated crypto sections already.
Another example would be the function of the insurance pools, or loss mitigation measures that work well preventing or during or after an attack.