Should be possible to get this to species level, but I'm not confident to do so myself.
According to Porter's guide to caterpillar
Five-spot burnet has "the anterior of the pair much larger and subquadrate"
Six-spot burnet has "the posterior of the paid smaller and more elongate"
I'm sure the author knew what he meant but as I read these they are both saying that the pair towards the head is bigger and squarer than the pair towards the anal end. Not sure I can tell them apart from this description.
Looking at the pictures in that volume I am coming to the view five-spot burnet is best option. Will add that and see what people think.
Observations here in Glos., suggest that 5-spot cocoons (and hence the per-cocoon larvae) are difficult to find as they occur low down on stems of vegetation. Difficult to see here what height the cocoon was found. Despite the diagnostic but not necessarily intelligible differences that you quote, it does seem to be a very late date for a 5-spot to start metamorphosis.
I don't think this was a cocoon but a parasitised larval skin. I suspect the parasites burst out quite late in the life cycle but it may not have been looking to pupate.
Lat/Lng: 51.254588, -0.315486
OS grid ref: TQ176519