I took some output from &uloc and put it into Google Sheets, and made some charts. Shown is the number of source lines compared to number of unique lines over the course of some git commit history. The overlayed line shows the rate at which no. of source lines is increasing relative to unique lines, ranging from 0x to 4x. In the first pic you can see the rate is somewhat tame compared to the second pic, where there's some spikes going up to 3x - 4x, hinting that the code added at those commits was largely duplicate code.