"On the other hand, there were these students who live-LaTeXed math lectures on their laptops..."

I'd suggest very few could actually do that successfully. First, they'd have to be truly fluent touch typists—that is being able to keep their eyes fully focused on the lecturer and being fully fluent in LaTeX.

And I'd suggest that's a very tall order if say the lecture was on math or physics—say on a topic such as the Navier–Stokes equations. Not only would their eyes be focused on the lecturer and what he/she's writing on the blackboard but they'd have to be sufficiently fluent in LaTeX math to enter rows of partial differential equations on-the-fly and be completely wordperfect about it.

No doubt there will be a few exceptional people who can perform such a feat but I'd bet the number would be miniscule.

Edit: when I was taking lecture notes I'd often resort to transcribing lectures from recordings to get the stuff that I missed or that wasn't clear at the time of the lecture. Unfortunately, back then there were very few video recordings of lectures which meant that transcribing math, physics etc. was essentially out of the question.

There were however audio cassette recording of subjects where visuals were less important such as philosophy. Even with these subjects I found that transcribing the audio into coherent notes was timeconsuming and difficult and often it required me to replay sections of the tape over and over. This was worthwhile as it drove the lectures home but it was damn hard work.