While that sounds reasonable, I wouldn't bet on it. I have seen too many inane bloat that was implemented simply because the designer/developer/project manager thought it was a cool thing to have or because they read it on some growth-hack thoughtpiece somewhere.

Indeed, that's all too common. It could also be a fluid thing. Huge images and lots of interactivity are in now, but it'll go too far (arguably has) and people may want to swing the pendulum back.

Very often institutions decide stupidly, when the decision is made by people that are not domain experts. Also, numbers may lie. If they did not study the same content with interactive and plain presentation, for example. I recall an article on a murder in Iceland on BBC IIRC, it was overtly interactive, every movement on the touch pad causing unpredictable changes to the page, but the content was so interesting that I put up with it.

Without being able to see what research they actually did/scruitinize their methodology, I don't find the claim that they checked very compelling. Humans are crazy-hard to study -- research that's not even peer reviewed is basically worthless.