> The triune brain hypothesis doesn't seem "wrong", it just seems like a simple categorization that is useful for making big-picture distinctions.
Right, but having worked in the field, I can assure you that there are, in fact, too many practicing psychologists, cognitive scientists, and even neuroscientists who believe the triune model of the brain to be literally and, sometimes, absolutely true. These are the types of people whom the linked paper is trying to reach.
Misunderstandings based on these oversimplified models are driving the current debate around modular versus distributed computation in the brain[1]. Obviously, a more accurate model of the brain would account for both ideas, but there is growing concern in the neuroscientific community over the amount of grant money going toward defending older, dead-end modular models instead of improving newer, more promising distributed models, mostly as a result of entrenched interests prioritizing the maintenance of prestige over the pursuit of truth.
In short, putting bad models on blast is good and necessary for the advancement of science. You can get a lot done with the plum pudding model[2] of the atom, but you can get far more done with the Bohr model[3] which emerged only after Rutherford, Bohr, and several other physicists published several iterative takedowns of the former, and yes, they too had to deal with entrenched interests who operated under the assumption that the plum pudding model was literally and absolutely true. It took a decade of experiments and several increasingly correct models before academic consensus shifted enough to accept the existence of subatomic particles and academic consensus began its collective investment in quantum mechanics. We're now in a similar place with neuroscience in the tension between modular computational models, which includes the triune brain model, and distributed computational models, which are showing promise in rescuing fMRI studies with their strong modular tradition from the replication crisis[4].
[1] https://www.cell.com/trends/cognitive-sciences/fulltext/S136...
[2] https://en.wikipedia.org/wiki/Plum_pudding_model