This sounds right.

When looking at the people who started scientific revolutions, it is the middle-aged, not young, who are overrepresented.

It also needs to be noted that during the last couple of hundred years, the amount of scientists in the world has been constantly increasing. The net result has been that there have always been more young researchers than old researchers, since more members of the younger generations have chosen to become scientists than happened in the previous generations. This has led to an illusion of youth being a requisite for scientific discovery, since there have been more young scientists and therefore also more young scientist geniuses than old scientist geniuses.

Scientific performance, as measured by the number of publications and the frequency of citations for those publications, increases steadily over time and reaches its high point around age 40 at least in chemistry, geology, mathematics, physics, psychology and sociology.

References:

Cole, S. (1979) Age and Scientific Performance. The American Journal of Sociology, vol. 84, no. 4, 985-977.

Wray, K.B. (2003) Is Science Really a Young Man's Game? Social Studies of Science, vol. 33, no. 1, 137-149.

Also, I seem to have lost the reference, but I recall seeing studies claiming that at least in academia, your creativity does drop as you age - but this is a function of career age, not chronological age. In other words, once you've been in a field for a long time, you stop having new insights. If you switch to a new field, you can start innovating again.

and (2) continuing to work on the highest level would require them to study the ideas of (and thereby subordinate themselves to) lower-status younger folk.

It's not necessarily the low status, but the fact that spending effort to study an idea is an investment, and an old person will get to enjoy that investment for a shorter time. It's apparently a relatively standard idea in economics that as people get older and their expected remaining lifespan shortens, they will stop investing as much in learning new things, since they'll have a smaller payoff from them. Richard Posner writes in Aging and Old Age:

One way to distinguish empirically between aging effects and proximity-to-death effects would be to compare, with respect to choice of occupation, investment, education, leisure activities, and other activities, elderly people on the one hand with young or middle-aged people who have truncated life expectancies but are in apparent good health, on the other. For example, a person newly infected with the AIDS virus (HIV) has roughly the same life expectancy as a 65-year-old and is unlikely to have, as yet, significant symptoms. The conventional human-capital model implies that, after correction for differences in income and for other differences between such persons and elderly persons who have the same life expectancy (a big difference is that the former will not have pension entitlements to fall back upon), the behavior of the two groups will be similar. It does appear to be similar, so far as investing in human capital is concerned; the truncation of the payback period causes disinvestment. And there is a high suicide rate among HIV-infected persons (even before they have reached the point in the progression of the disease at which they are classified as persons with AIDS), just as there is, as we shall see in chapter 6, among elderly persons.

Later on, he also notes that various careers vary in when they reach their peak:

The first thing to note is that the very concept of a peak age of productivity is misleading in suggesting that all careers have a sharp peak. There are careers with early peaks and careers with late ones, but also careers in which the peak, whenever attained, is sustained without a significant decline virtually till death. Let us call these "sustained peek" careers, as distinct from "early peak" careers and "late peak careers". Sustained-peak careers can in turn be divided into "early peak, sustained" and "late peak, sustained", this giving us a four-fold division: early peak, not sustained; early peak, sustained; late peak, not sustained; late peak, sustained. Examples of the first category (early peak, not sustained) are most fields of professional athletics, along with mathematics, theoretical physics, chess, heavy manual labor and - the analysis in chapter 6 implied - most criminal "careers". In the case of physically demanding activities, risk of injury plays a role; it is more difficult to sustain peak performance in football than in dance.

Examples of the second category (early peak, sustained) are literature, economics (other than the severely mathematical), musical composition (including choreography), painting and sculpture (consider Michelangelo, Titian, Picasso, and O'Keefe, among others), and musical performance. An example of the third category (late peak, not sustained) is the senior management of large firms, where the peak age will often be in the late fifties, followed by retirement in the early sixties; perhaps most leadership is in this category. The fourth category (late peak, sustained) is illustrated by judging, discussed in the next chapter. History, theology, literary criticism and scholarship, and philosophy appear to straddle the second (early peak, sustained) and fourth (late peak, sustained) categories.