Much less people have the aptitude for quantum physics which is way harder. Programming is so easy that even as a teenager you have sufficient background to learn it.
I assure you that as a teenager, it's very unlikely you had any aptitude for quantum physics or any of the hard mathematics required to understand general relativity.
I find doing advanced math relaxing and meditative during commutes. That's aptitude. It's probably much the same aptitude as finding programming enjoyable, but software is a much better career, so most of us end up there than in physics anymore.
Not many people have this aptitude, though. Most people would name hundreds of activities they'd rather do than either of them.
But let me put it this way. Much much much much much More people have done programming since they were teens but much less have done QM. You will find tons of examples of programmers and relatively few people who know QM or advanced mathematics.
That's the dichotomy I'm talking about. Programming is easy. QM requiring pre-requisite knowledge is part of what makes it so much harder.
You analogy is bad. But I get your point.
Let's put it this way. Something that is commonly done is LIKELY to be far more easier than something that is not commonly done. Full circle back to our other point on probable causes.
This is obvious. Your diving into technicalities of language, which is unnecessary.
My claim is that most of the population can achieve the average skill level held by an average developer.
I also wouldn't call it a "selectivity theory." It's more common sense. You select a group of people with a record of great past performance they are more likely going to have an easier time. It's actually your ideas about selectivity are counter-intuitive.
Programming is what I did obsessively as a teenager. I didn't need any discipline or study habits. It was and still is exciting and creative. This is aptitude and interest, which obviously, few have.