Thanks Tron. The sharpness was basically human error. I may be being a bit hypercritical of my attempts, but I was trying to run before I could walk. I was trying to focus at the hyperfocal distance (10m @ 24mm and f/1.4) to get some of the foreground in focus as well, when I should have gone for the shot first, then started experimenting. One of the criticisms of modern lenses (and it is one area where Zeiss lenses have the edge), has been the depth of field scales. Basically, the scale on the 24mm f/1.4 MkII (and in fact most Canon WA lenses) goes from 3m to infinity, making it a guestimation that it is focused at the hyperfocal distance, unless you prefocus on something at 10m. I could have afforded to stop down a bit too (which I did later on, when the lights had subsided a bit).
The ISO 800 shots were 15 secs, so it is likely that there would have been some star trails, while the ISO 1600 shots were at 8 secs (mostly f/1.4, but some later ones at f/1.6). The lights weren't exactly bright, but at least I saw them and got some usable shots. Incidentally, the two shots that I thought were teh best in terms of sharpness (both when the lights were subsiding), can't have been too bad, as they were accepted by Alamy, in a batch only containing the two shots and nothing else. Alamy typically scan two images from each upload to make sure the image quality is high enough.