I don't argue that there aren't such cases, my point is that it's very luck based and it's basically a lottery. You might as well start streaming on Twitch because it's the same kind of lottery where if you make it, you make it.
You listed a few projects and I can add a dozen more on top of them, but statistically those are probably <0.1% of all _finished_ startups led by 1-3 people.
I have worked in startup business as a senior full stack web developer, and I've met many people there who just hop from one startup job to another, because startups even if properly funded just never make it. You make a big buck as a developer there because of all the risks, plus they often offer you a share after the product has gotten popular, but it never gets popular. That's just how IT business is.
The advice to "just make something that will give you money" faces the exact same problem. It's literally the same market and in it you're also at a disadvantage, let alone the amount of competition.
Sure you can make yet another basic app or something like that, but at the end of the day it's not even the size of the project, it's just how lucky you are in getting it running. It can be your 10 year long monstrous project or it can be a 6 hour hackaton challenge project. They both have equal odds of becoming big enough to bring revenue.
This is just an unreasonable investment of time and money, is my point. If you like gambling then it's probably ok, you can keep trying to push your ideas and develop more stuff. But if you're into having a stable financial situation then it's probably not for you. I consider myself to be having the preference of the latter.