I think that Knuth's concrete mathematics made this connection at some point - the difference between the infinitely small abstraction of an integral and a big-sigma sum as being discreet quantities that can be measured by actual numbers that measure a thing.
