Human follicular development

I teach the human ovarian cycle in my class as an elegant example of what happens when you have multiple hormones with complex, yet understandable, interactions. It's a neat, fairly well understood system, so you can imagine my interest when I found an older entry on Nature Is Profligate citing a recent study showing that the cycle doesn't work quite as I understood.

Women's eggs are stored as undeveloped follicles that mature during the ovarian cycle. Follicular development is typically described as a process wherein one follicle leisurely matures, and once it reaches maturation, ovulation occurs (with a number of hormones playing a role). More detailed textbook descriptions often include that multiple follicles are stimulated to develop at the beginning of the cycle, but that after a week or so all but one of the follicles typically regress. In both cases, however, the process is usually presented as a steady progression from an undeveloped follicle to a mature follicle that is ready to ovulate.

However, it appears that women almost always have multiple waves of follicular development. A recent study (see both the paper's abstract and the journal's summary of the study) used ultrasonography on 50 healthy, regularly menstruating women to examine their follicular development (sidenote: that's some detailed ultrasound!). They found that 68% of the women had two waves of follicular development, and the rest of the women (32%) had three waves of follicular development. Based on their data it looks like no women had a single wave of follicular development. It appears that most women ovulated only once, meaning that the follicles in the first wave began development and then regressed, allowing another set of follicles to develop afterwards.

So how could we explain these multiple waves of development using classical hormone models? A rise in follicle stimulating hormone (FSH) causes follicles to start developing, and the developing follicles produce estrogen which then reduces the levels of FSH by negative feedback (until just before ovulation, but we'll ignore that). On first glance it seems that we can understand how a 2nd wave would start after a follicle regresses: the regressed follicle stops producing estrogen, which releases inhibition of FSH, leading to a subsequent rise in FSH and thus stimulating another follicle to develop. (Note: I'm extrapolating here, and I've removed gonadotropin releasing hormone and luteinizing hormone from this discussion to simplify things)

But my question is this: what causes the waves of follicles to regress in the first place? Would that be under hormonal control, cellular control, or some combination? Any thoughts? Any other neat work on this topic?

Have I just been the victim of oversimplified textbook descriptions all along?

The paper referenced above is: Baerwald, A. R., G. P. Adams, and R. A. Pierson. 2003. A new model for ovarian follicular development during the human menstrual cycle. Fertility and Sterility 80:1 pp. 116-122.