Identifying one edge case doesn't really prove an argument. Net wealth is not a perfect metric, but is probably the best one we have.
You can be sure 95%+ of brain surgeons have paid off all their student debts and have a lot more wealth than a homeless guy, considering their median salary is $395k per year.
Those wealth statistics really are dodgy. Another edge case: you're earning so little that you don't save, but spend it all on food/housing/etc. Now your country's economy takes off and you suddenly earn 20% more. You're still spending it all, enjoying much better and healthier food. To those statistics, nothing changed.
Well, show me a perfect metric that doesn't have similar edge cases?
For example, income has its own issues -- a retiree can have little income but a big retirement fund to draw down.
In an ideal world, you'd probably measure something like current net wealth + discounted expected future earnings, and maybe use PPP instead of GDP, except now you're trying to measure two things accurately instead of one, and making guesses about the future too.
Measuring wealth is a pretty poor way to measure inequality.
If you're a recently graduated brain surgeon with $400K in student debt, you have less wealth than a homeless person with a $100 bill in their pocket.