>
angular power spectrum of the CMB
My bad. I'm not familiar with this.
(Beware, my English is weak).
I had year long discussion about similar topic by email, so I see nothing exceptional there. If you have sum(sin(x_natscale)), and you will start to play with scale, you will see sinusoidal graphic, freq=asin(scale). Object formations in space usually have round shape, so it's will be same in 3D. :-/
Why? Because our formation exists for very long time. Let's play simple example: we need a 1d function f(x), such as sum(f(x_nat*scale)) newer produces infinite sum or singularity. Physical meaning: Our Universe exists for infinite time, but all matter is not attracted into one single point of infinite mass, so it's not possible.
The simplest such function is sin(x) or cos(x). If we will try to play with scale, we newer be able to make infinite sum, because Pi is irrational number. The closer our scale will be to Pi, the larger peaks of sum we will have.
In real life, this influence diminishes after some point, so, for example, size of nuclei has influence at frequency (distribution) of chemical elements, but it has no influence at macro objects, so function becomes flat and sum must rise (e.g. collapse into a black hole), then next level begins.
So, it looks like at some point, these formations cannot be bigger, or we cannot see light from them, e.g. because they are so massive, so light cannot escape them. If so, we have lower limit of formation size we can see at this distance, dictated by geometry (angular size) and luminosity, and upper limit dictated by nature, e.g. upper mass of formation of such size (upper density).
It's like in our galaxy: we can see stars, but cannot see black holes (too heavy), lone planets and smaller objects (too cold), and far away objects after certain distance (too small luminosity). IMHO, angular power spectrum of visible stars must show distribution similar to CMB or distribution of chemical elements.