While free throws is an important part of the game, it's relatively minor compared to other factors that make basketball players great - athleticism, basketball IQ, ball handling, height, playing under pressure, defensive ability, etc..

I wouldn't be surprised if other metrics have been steadily increasing over the past 50 years. Basketball players in college and the NBA simply aren't selected for their free throw ability.

The fact that is has been approximately constant over such a long period of time is what I find most surprising. I would naively expect many of the improving factors that you mentioned would lead to an increase in free throw percentages as well. Clearly, there are some teams which hit >75% and even occasionally close to 80%. I like the idea that there is a tradeoff in that every hour spent training for free throws leads to a decline in other skills. But still what's special about the 69% number? I would like to see how much time the 75% teams spend doing free throw drills compared to their competitors.

Similar to optimisation theory, each skill has positive and negative effects on other skills. This has been thoroughly explored in fields such as profit maximisation and breeding, but little has been done in sports.

Why don't they just use the granny shot? The cross section of the basket is much higher from the typical granny-shot trajectory is it not?

It is, but it is still an inferior technique for two reasons:

1. It is impractical as a technique at any other time because it is so easily defended. This means that it would need to be developed as an additional skill.

2. It requires the coordination of both hands. Proper shooting form requires only the use of one hand and the elbow is in line with the basket. Good shooters generally only miss on distance for this reason.

I think the takeaway from the article is that a free through is bound by players' psychology rather than athletic ability. Other accuracy based solo-efforts also haven't shown improvement:

There are measures in other sports that have shown similar consistency, like golf scores or batting averages, but none of them are as straightforward as lobbing a ball toward a basket.