I agree. It looks pretty random. Its probably even more random since we can't reliably eyeball the chirality in most of photographed galaxies so half the angles in one of the axis are practically mirrored at first glance.

That observation in itself is not a disproof.

(Imagine standing in a room with a bunch of CDs suspended horizontally from the ceiling. From any vantage point you can see some CDs edge-on, some from the top, some from the bottom, etc. But they're all oriented in the same direction.)

This situation is trivially detectable though. Simply plot the inclination angle of disc galaxies as a function of azimuth and inclination, if it's like you say, there will be a very clear banded structure.

That analogy does not hold up when the CD you are observing from is in the same plane as all the others.

I suspect that in your example you would observe a number of different apparent orientations, but that there would be some types of apparent orientation that could never occur.

Ok, my mind is boggled too... Anything we can read on the topic ?

What topic? Basic geometry?

I feel like they might have noticed if it was that simple.