I have observed that when I pulse cells with EdU for analysing cell cycle kinetics, the pattern of staining is as in Fig.1. Considering the heterogeneity of S phase cells during the pulse, cells that are in early S phase appear as EdU[medium] DNA[2n] (marked by the blue box); cells that progress through S phase appear as EdU[high] DNA[2n-4n] (green box). However, cells that are in the late S phase must appear as EdU[med] DNA[4n], which I do not see (red box). I have pulsed cells for different durations (2h, 24h, 48h) and still observe similar distributions. I use propidium iodide for measuring DNA content.
Fig.2 is from the manufacturer's website. BrdU images from publications also show similar trends.