Patrick Chase
It makes perfect sense if you've actually looked at it in any depth.

Monocrystalline diamonds (and CBN) do wear at the tips. Look closely at this monocrystalline particle from a broken-in DMT #325 for example:

https://scienceofsharp.files.wordpress.com/2015/09/dmt325_brokenin_03.jpg

There is a "before" shot in that blog post as well, but the short version is that the exposed corners do break down, and quite dramatically. That's one of the causes of the dreaded "diamond plate slow-down effect". CBN is known to be worse in that regard than diamond, due to its lower hardness (about 2/3 the hardness when comparing synthetic monocrystalline diamond to synthetic monocrystalline CBN).

They key thing to realize is that "monocrystalline" doesn't imply a single perfect diamond lattice. It just means that the bonds between regions are metallic (crystalline). Synthetic monocrystalline industrial diamonds are both softer and more friable than natural ones, though far less so than resin-bonded polycrystalline ones.

You can get some idea of the factors in play by reading manufacturers' literature, for example: https://www.hyperion.sandvik.com/en/products/diamond-and-cbn/diamond/micron-diamond-cbn/diamond-micron-powder/metal-bond-diamond-micron-powder/. All 3 grades there are nominally "monocrystalline" (metal-bond) but differ appreciably in cost and hardness. The compounds I use are "SJK" grade, and I think that most plates on the market are similar.

If you look around you can find suppliers like Precision Surfaces and McMaster selling natural diamond powders for lapping super-hard materials like solid carbide. The reason is that economical synthetic diamonds simply aren't up to the task.

So with those details out of the way, what I said stands: Diamond crystals do wear out by breaking down. CBN is worse in that respect than diamond.