Just a thought from an engineering perspective. In my experience designing electronic gear using digital gain controls, glitches or spikes at the stepping points (zipper noise) are often exacerbated by any DC offset in the system. By that, I mean a small fixed voltage which is always present in addition to the varying AC voltage signal. The actual offset is generally very small (so isn't enough to cause any harm such as asymmetrical clipping for example). It can vary from device to device and is generally due to internal component tolerances so is difficult to characterize. In higher spec equipment, the DC is generally nulled out, either automatically or via adjustments in the test phase, but this is probably not cost effective with tiny devices like the R7. So I would guess that that some R7s , or some batches, produce more audible zipper noise than others.
Not that any of this helps anyone who's experiencing the problem! But it might explain the variation which people are reporting here...