This is a good article, but it only touches on the main issue with learning styles. Multiple independent groups of educational psychologists (if I remember correctly it was three) reviewed all the articles they could find on learning styles. They found that the majority of the studies had poor experimental design, since they were missing proper control groups. Of the remaining studies there were more examples of negative results for learning styles than positive results. This was pretty damning, given the publication bias for positive results in academic research. The article links to one of these reviews, but here is another one:
https://www.psychologicalscience.org/journals/pspi/PSPI_9_3....
I teach science, and it's frustrating that nearly all of my students can tell me their learning style. Their previous teachers have taught them about learning styles as a way to improve their studying, and they've jumped on that information. But the literature indicates that matching instruction to a learning style doesn't improve learning. This is a total waste because we're effectively giving students a magic talisman to help them learn science, because once students find out it's wrong they'll be less likely to take our advice on studying, and because the time we waste on it could be spent on telling people about real things that actually improve learning.
The studies I have seen (I'll admit- I didn't read TFA- maybe I have in the past idk) all used some sort of survey to identify learning styles. Nobody, to my knowledge, has measured learning by modalities to identify them.
You are spot on that attempts to match instruction to learning style have been futile in the VARK-like learning style classifications. They also assume that if a student has a learning style, it will be the same across subjects. If it turns out there is such a thing as learning styles I woudn't be at all surprised if it varied by what you are learning.
Just as a thought, just because it doesn't improve learning outcomes doesn't mean it doesn't improve the learning experience. There might be a qualitative aspect missing from dismissing learning styles. Even if learning styles only improve experience as a placebo/illusion of self-control.
I teach science, and it's frustrating that nearly all of my students can tell me their learning style. Their previous teachers have taught them about learning styles as a way to improve their studying, and they've jumped on that information. But the literature indicates that matching instruction to a learning style doesn't improve learning. This is a total waste because we're effectively giving students a magic talisman to help them learn science, because once students find out it's wrong they'll be less likely to take our advice on studying, and because the time we waste on it could be spent on telling people about real things that actually improve learning.