I wonder even if these percentages also change in time for one person. And if people are spending more time with something in ways they don't realize, which kind of distorts recommendations.
I noticed in other subjects, that while students start out most busy learning fundamentals at school, in time the focus changes more to the analysis of other people's work by those fundamentals in university. And what of these higher works are in the realm of fundamentals, and what in another's style, becomes a bit diffuse even. Because style derives from fundamentals, there is always a little to be learned even about fundamentals in just style.
There seems a point of saturation in most subjects, where people concentrate more on a sort of style specialization, even though fundamentals remain important, and are even subject to ongoing research still, by others that keep concentrating on them instead. But usually that isn't to say either is wrong in their different emphasis and goals. Rather even synergetic.
So from a more general perpective, looking at universal problems of life, it keeps making me wonder.
On the other hand, I also do agree a lot, that I often think that most people in the field of programming too are not well enough educated on the fundamentals of computing, and that this hurts them badly in their higher level decisions very often. Usually, they will argue in favour of the quick results they are getting by relying on other people's style interpretation of computing, in languages/frameworks/engines, results they seem content with in the scope of their project. And even though it has so many obvious flaws, unwitting misuse, and unnecessary limitations, hidden behind a shiny surface, maybe they are still right, even in that other things matter more to most people using it. But it can be frustrating of course, to watch them over-do their justification, that nothing else would make sense than living these abstractions. Why they shouldn't aim for more, for going deeper. Why others shouldn't waste their time trying. Often stating strong opinions as fact, while at best relying on hear-say, without ever having done themselves work on compiler construction and language design to qualify their statements. Sometimes I think, a coder isn't a coder, unless coding in his own language. Still, I try to see more the creative opportunity they have in not having to do all that, what aspect makes their work worthwhile, and how it may not really work well otherwise the way it does, all things considered, that they'd miss other opportunities in bogging themselves down too much with fundamentals. Maybe. Or maybe, people really do not spend enough time in the fundamentals of anything, and we should be more consistent in our critique.