Agreed, it's quite believable that the threshold of capital required to develop a given improvement is increasing. That's one reason that individual entrepreneurs are likely making fewer impacts on starting of new companies. Well that and there is a double whammy of labor being paid less of a margin of productivity so individuals and even small businesses are more squeezed of their own capital accumulation. In other words, the Wright brothers today would have a hard time not only because it's harder to develop a competitive aircraft, but because their bicycle business isn't going to yield a sufficient margin to even invest in the development in wind tunnels, motors, etc that they did originally. The thing is, it's individuals that would make the 'crazy' investment without a cut-and-dry ROI, without a calculable guaranteed result - and I think we suffer from their absence in terms of end growth overall.
But I also somewhat discount that the rising difficulty level is really why more breakthroughs aren't happening, if that were the case one would see larger and larger accumulations of ventures committing capital and failing (or at least operating). I think the real reason is that not enough capital even tries to attack these problems because they're longer term and difficult to quantify the return. The irony is that I suspect that we get a smaller overall growth by routing capital to mostly knowable investment returns over a larger proportion going to "crazy" non-calculable investments. So, we end up waiting for the die-rolls to match up people like Elon Musk with sufficient fortunes for truly big advancements, and an economic system that mucks around with financial engineering instead of relentlessly and systematically advancing a wide range of real-world applied engineering on big problems.