The systems that I design always include large arrays of data acqusistion channels, implemented with an assortment of communication protocols, all running at different speeds. I have always wondered: Has anyone ever thought of or estimated the approximate "speed" of human speech and its comprehention? For example, 30-baud. Is the variability of human speech even quantifiable? Thanks for considering my "goofy" question. Perhaps there might be a more appropriate forum within which to ask my question.
I think this has indeed been considered and a keyword to find related work would be the "entropy" of natural languages. See for instance this paper https://arxiv.org/pdf/1606.06996.pdf obtained on a search engine, I am sure many more exist.
(I am also of the opinion that, unfortunately, this question is indeed not really in scope for TCS.SE.)