I decided to pull down one of my SP rounds since it was so grotty anyway. For comparison is a 115gr bullet from an HN load.
The base of the FMJ load (at right) has a slight crimp line around the tail of the jacket just adjacent to the lead. This line does not appear on the RG SP base.
So, I pulled down an RG 1956 round (at left) and—no crimp line.
Below are the later stages in bullet manufacture from a Polte board from about 1934. It looks like the tail of the jacket is turned into hold the lead core in place in one stage and the next stage (different tooling) finishes the base of the bullet. In this case it appears this last step is done in two stages.
I have checked draw sets from four other countries and they all form bullets basically the same way. Unfortunately, none of these sets are RG or even British.
It appears that if the fourth or third stage from the right were turned upside down it would take a considerable increase in pressure to reshape the entire bullet, compared to what appears to being done in these last three stages. In addition, it appears that if the fourth or third stage were inserted upside down in the feed line they would not fit down more than about halfway into the next stage die and more likely cause the machinery to jam rather than totally reshape the bullet.
No data for a firm conclusion, but I am still having great difficulty figuring how these bullets could be made by mistake.
One possibility is that the SP bullets were made intentionally for Special Operators, but when RG switched from SP to FMJ, they didn’t always clear the bullet machines and as a result a few of the resulting SP loads got mixed into the normal FMJ production. This could also explain the Geco SP loads mixed with FMJ ammo. Again, just speculation.
A couple of hours work, and no better insights. I’m going to go upstairs and have a wee drop of Whisky from a land where they talk funny…