Indeed. I suspect reflections on the diaphragm blades to play a role here, but I have no explanation. Another way to look at it is to consider that flare always exists, and can be measured in terms of signal to noise ratio. On stopping down, the signal gets reduced by the diaphragm. Even if the noise coming from internal reflections would remain constant, the ratio would decrease.
Strangely, the Zeiss guru (see diglloyd.com) did not spot the issue, and on the contrary praises the flare-free character of the lens with one or two shots as evidence. On my side, I observed the phenomenon repeatedly, the above scene really being the "acid test" (from experience with many other lenses). I did not however inspect the full range of angles of incidence, etc. Be assured that in most cases, the flare is milder and/or localized rather than generalized. This lens remains my favourite - just not for shooting organs in churches, in which case I'd revert to the Zeiss 35/2, or Nikkor 50/1.8 G or 50/1.2 maybe, to remain around 50mm.
Coma is a different phenomenon, and a predictable aberration rather than a design flaw. The coma of the 50/2 MP is moderate, and no real issue for night photography. Also, since night shots are essentially underexposed (the night sky is expected to be dark, not 18% gray), and using apertures around f/2.5-f/4 typically, the flare issue is not acute. I have used it for such purpose. I guess the Milvus version is at least as good, and this is supported by JJChan's night shot.
In search of the universal 50mm, I could not find one but the Zeiss comes close to it. The Milvus might be even closer. Now let us see if the strange behaviour of the Zeiss MP is still observable here.