Ok, I'll take a swing at this one. It's just the nature of the beast - in machines of any type - the parts aren't perfect. There are clearances and tolerances, ranges of performance and fit specifications.
In this case, I think the key players are the oil pump, the shaft, and the distributor gear - cam gear interface.
The oil pump does not present a uniformly smooth load to the shaft. The twisting force required to turn it varies as the pump's internal lobes engage and disengage.
The shaft is not perfectly rigid - think of it as a long, stiff torsion spring. The non-uniform pump load makes it twist and untwist as it rotates. As it does so, it stores and releases energy, some of which is lost to friction.
Then you have the distributor gear - cam gear teeth, which have to have some clearance between them. Normally, the oil pump load will push the teeth into contact, and the whole shebang will rotate more or less smoothly.
Now, depending on the mass of the parts, the springiness of the shaft (and other parts), and the amount of friction, this system may have a frequency (engine speed) at which it will resonate - or ring like a bell. When it does, the shaft may be twisting and untwisting so much that the cam and distributor gear teeth are not in constant contact. One of the results is spark scatter.
Notice that I've simplified this model by ignoring the imperfect crank to cam drive, and the fact that the cam presents a very non-uniform load also. Not to mention the poor little roll pin that holds the whole flying circus together.
As is often the case, something that appears relatively simple is really quite complex when you take a good close look at it.
Anyway, I hope this helps with understanding the sometimes not so obvious behavior of our wonderful machines. Rodney