Well, OK, I'm not disputing that. But this discussion isn't about those high-value projects, it's about using fonts in native apps.
And if you don't think what you're building has the potential to be high-value, why are you working on it?
Because not every project in the world has to be Google or Facebook. I think your definition of high-value is probably a lot higher than what many people would consider a successful project, assuming the project is even being done for profit in the first place.
You don't seem to get it: it is a perfectly sane business decision to choose your client base.
I get that just fine. It's just that you seem to be the only person in this entire discussion who is talking almost exclusively about high-end projects, while the rest of us seem to be generally agreeing that the current state of type licensing is not helpful for anywhere below high-end.
The rest of us are talking about what is holding back typography in apps and how the current licensing regime isn't promoting the use of better fonts in that context. It's right there in the title of the discussion.
That said, there are hundreds of thousands of apps out there right now. Unless you think someone from each major foundry is checking every one of them to make sure they're not infringing, it is self-evident that there are very many companies so small that nobody will notice if they infringe the copyright on a font or two.
It happens all the time, when creating apps, in professional design work at studios, when creating content in-house, etc. Pretending otherwise is like pretending software doesn't get pirated by small businesses all the time, completely unrealistic. You might not have to be Google's scale to get noticed, but I expect the vast majority of people working professionally with fonts could use any extra ones they downloaded illegally with complete impunity. But every time they do, that's probably a failure of the market to generate some revenue for whoever created that font.