Einstein, Ramanujan, and Newton were boosted by existing networks of review and promotion. A lot of core engineering math was invented by aristocrats and government functionaries around the French Academy. Germany developed its own equivalent scene somewhat later.
All of these followed the model of a relatively small number of smart people bouncing ideas off each other, reviewing them, building on them, and promoting the good ones.
The difference between that and modern R&D is that modern R&D tries to be industrial rather than academic. Academia is trapped in a bullshit job make-work cycle, where quantity gets more rewards than quality and creativity. There isn't room for mavericks like Einstein. Even if they're out there having great ideas, there's no way for them to be discovered and promoted.
Industry focuses more on fill-in developments than game changer mathematical insights, which are the real drivers of scientific progress.
So there's a lot of R&D-like activity in CS, and occasionally something interesting falls out, like LLMs. But fundamental physics has stagnated.
One of the biggest reasons is that the smartest people don't work in research. They work in finance, developing gambling algorithms.
I'm not discounting the benefit of having additional networks in place, I just think they're a facilitating factor rather than a causative one. They're important for educational development and spreading ideas, but they can also result in homogeneity. The biggest two factors to me are time and interest. You yourself point out that most modern math and engineering was invented by aristocrats; the main reason for this was that they were the only ones with the luxury of being able to think about such abstract topics. They had the time to spend, so those with the interest and aptitude pushed the boundaries. You occasionally saw members of the working class do the same (Heaviside, for example), but they had a more difficult road.
I disagree that the smartest people work in finance. Some very smart people do. From what I've seen, the ones at the very farthest edge of human ability typically aren't motivated by money.
If for example you knew someone who has great idea, but doesn't have time to check it with a prototype because he has a lot of "normal" work, but he doesn't want to give it for free to everyone (so that he can finally have some more money and check his other ideas), what would you suggest to that person, or how would you support such person?