As I recall, roll #1 and roll #2 were quite close although in hindsight things had started drifting a bit already on #2, and #3 was distinctly different. There was also a color shift, but what surprised me was the difference in acutance as I couldn't explain that. It was not a structured test BTW, so that made it difficult to analyze.
Ok, some of the change you noticed would probably be decreased by a slight change in development time as suggested by EK in their posted chart. That is just OTOMH.
Yes, that makes sense. I decided to go for one shot though, to get rid of this sort of variance. I had expected it to be less, so I was kind of surprised at the outcome.