Richard: thank you for revealing the term “capsule aperture” to me. I may well contact Russ and suggest that he consider adding them to APT. He has been very welcoming of suggestions for enhancement in the past.
For only slightly trailed images (major/minor axis < 2, say) an elliptical aperture works very well – as measured by comparing the results for untrailed subs and circular apertures with their trailed counterparts of the same field and a carefully chosen elliptical aperture. If you examine some of my entries in the VSS database you will find some examples of elliptical aperture photometry.