The bad news is that a lot of people will never attain the formal operational stage, and what is even worse is that the percentage of people that do is falling.² The results of this study are not surprising.
The take-away should be that it is pointless and counter-productive to try to dumb down the user interface for applications in domains that require a high level of cognitive development. No amount of picture icons and hamburger menus is going to make up for an inability to reason abstractly. What it will do is make it impossible for anyone to ever use the application efficiently.
¹ https://en.wikipedia.org/wiki/Piaget%27s_theory_of_cognitive... ² https://oneofus.la/files/flynn2018.pdf
I guess the equivalent in computing would be spreadsheets? You can automate and build awesome things but god help you if you wanna move it somewhere else
Similarly with spreadsheets. Sure, they get ugly. But they allow to get work done better and faster that it would be otherwise (if it was done at all). The alternative usually involves finding a ready-made domain-specific application and adjusting your (or your company's) workflow to it, or pestering someone to make it for you, both of which are frequently bad trade-offs. It takes time to arrange, often siloes you off your data, you have lots of irrelevant stuff to learn besides your task, and it prevents you from iterating on your workflow. I'd say the last one - iteration on workflow - is the biggest source of Excel (ab)use. Hard to lock down your process in a domain-specific tool if the details of that process are fuzzy and subject to change.
Computer programming is like building, modding, or repairing your car — you can learn but it’s an activity done basically only by hobbyists and professionals.
Troubleshooting computer issues is like car maintenance. Everyone can fill the tank with gas, most people can change a tire, fewer can change their oil, and fewer still can jump a battery.