The ability to discern the seam is a function of many factors. Those would include the contrast between the different glass colors, the shape and sharpness of the cutoff mechanism, the stream speed of the hot glass at cut-off, the opacity of the glass colors, the subsequent machine forming action after cutoff, glass temperature, cooling rate and time to cool to sphere shape in the rounding mechanism (longer = probably better smoothing/blending between stream colors). Each one of these are variables which can (and did) change from day to day and through wear on various key machine parts.
Having spent a lot of time at the Akro plant site, examining a lot of marble rejects, and owning some Akro machine parts... I got some sense of what went right and how things looked when they went wrong. The scale and depth and number of marble reject dumps were HUGE. One of the first things one sees is that marble production was a very low precision, fairly variable volume business. I got a sense what the machine did at the beginning of a run, what they looked like when things got "dialed in" and what the end of run and malfunction pieces looked like. The variability of the glass temperature and the glass colorant mix was greater than one would think. Marble production was not rocket science - by a long shot.
Seams varied due to these many factors (and probably more) - so the ability to typify one company's seam is IMO very difficult and assumes constancy of manufacturing method and process over time.