As I see it, an essential point of "Data Science" is to finally introduce mathematical awareness to the world of programming. I welcome our newly arriving & refreshingly numerate overlords!
"One way to read the history of business in the twentieth century is a series of transformations whereby industries that 'didn't need math' suddenly found themselves critically depending on it." This is a significant understatement. It also summarizes the history of business in the 19th century and even earlier. See, for example: steam power, metallurgy, bridge building, chemical industries, and electricity.
Not all hackers miss Miller's point. One of Zed Shaw's older rants "Programmers Need To Learn Statistics Or I Will Kill Them All" comes to mind[1].
Perhaps I'm missing some intended irony, but Miller's polemics on Lisp and his reverence for Fortran feel overdone & unnecessarily narrow. Coders who buy into Miller's notions can join the "Fortran School" by learning something along the lines of R programming (lots of fun, functional, array-based, and adept at statistics) or J (compression and austere beauty that only a mathematician could love) or any number of other languages that have not shunned math. Of course, calculus, statistics, and linear algebra also required. For the university trained, brush up (or curse your university for not requiring you to learn these subjects). For the autodidacts among us, the likes of Coursera and Sal Khan are indispensable.
[1] http://zedshaw.com/essays/programmer_stats.html