The feedback on my first attempt at turning a buffer suggested that the problems with the surface finish were related to the speed at which they were being made. Specifically that I was probably turning the lead screw too quickly, but also that I might need to run the lathe faster than the 1500rpm I was using. I've now had another go using a faster lathe speed of 2450rpm while turning the lead screw much more slowly. This reduced the problem considerably although I notieced that the grooving was still present but was only happening when reversing the lead screw so was solvable by backing off the crossslide after taking the last cut to set a specific diameter. After a few false starts, mostly related to me not being able to count, I've now produced four almost identical buffers that I'm more than happy with.
In total I've turned 10 buffers giving me a success rate of just 40% but for my first lathe based project I don't think that's too bad at all. One thing I did learn is that I really want some resettable handwheels. In the end I resorted to facing off the brass bar until the dial read 0 so that it was easy to count out the other measurements along the bar. I also know I really want a collet chuck for holding small items as doing the final work with the buffer held in the 3 jaw chuck by the locating pin was fraught with problems and in one case I ended up with it clamped out of centre somehow which ruined the buffer when I added an off centre notch around the front face.
Anyway, I'm really happy with how these have turned out and in retrospect I'm actually quite impressed that the success rate is as high as 40%, I thought I'd go through a lot more brass than that before producing four usable buffers!