You have 4 discrete die pairs with the oval reverse shield (3B-3E). The obverses are also different in each case, but still given a single number based on the design with the tail behind the rear legs. Obverse 2 has the tail between the legs and is known for 12 discrete die pairs. Again, the obverses and reverses are slightly different in each case. The flaws appear to be where a previous die was engraved as you can discern details which agree with other dies. It is this that leads me to believe that the cylinder press theory proposed by Besly in the 1984 BNJ doesn't hold, or at least only for the obverse 2 shillings and possibly the threepences. The above and the other halfcrowns are individual die pairs used on a rocker press, and are struck from diestock that is of a greater diameter than that of a halfcrown.