Thanks for the log, almcl!
I ran through it quickly as a test, seems to match up well with what you posted and was used in the blinky gif. But, with the ability to zoom to 400% or better, I'd say there's still what I would deem unacceptable ringing. That may wash out and not be very perceptable once we get to Color, but I didn't go that far.
Some differences. I loaded as bicolor rather than straight RGB, considering that more appropriate to the acquisition. But really that should just change the synth L by something like a 1.44% G bump. Noisier? Maybe. I also had stretched no-ROI, but the ROI here is fairly mild. I did use Contrast (can't recall if defaults or if I softened it by lowering the locality slider) but HDR only mildly, not defaults which I felt a bit strong here. Also I skipped Sharp.
In SVD I copied and then loaded your 5 PSF's. At the outset, the appearance is what we have been seeing of late. At least with my data, Stefan's, maybe Ron's. And again interestingly, some "other people" data I try works out much differently. Including a buddy with a RASA11 and dark skies, and he always takes like 2 minute full spectrum OSC subs and blows out countless star cores lol. Go figure.
I didn't fiddle too much (I should be on the freeway heading to the office lol) but just tried your 13% fuzz, which helped. The focus made little to no visible change on pre-post tweak toggle after that.
Now, some of the variance in results could be the stretch and the use of Contrast, HDR, and Sharp such that certain affected stars weren't surrounded by as bright a background, but I'm not entirely sure. More experimenting needed later.
By toggling back to the sampling screen, I maybe (?) noticed what appeared to be a distinct difference in the deringing effect between stars that are masked (though not necessarily selected) and those that aren't. This may explain some issues I've raised before as it seemed to be hinting at some sort of dichotomy - i.e. vast differences between the deringing controls as to different stars. I had wondered if there was a background support mask, but perhaps it's still just the sampling mask (don't know if we still call it an apod anymore)?
Anyway, masked stars don't seem to need as much deringing settings as non-masked. Hence, you can get side-by-side stars that behave completely differently, even if they seem fairly similar otherwise. And one can also get tons of little tiny stars with weak or no deringing, requiring some custom pixel clicking in Shrink later.
EDIT: On further thought, I suppose there could also be something about the star characteristics themselves that led to some being in the sampling mask while others were rejects?
Or I'm just completely off my rocker.
I'll go back at it when I have more time later.
And I thought I only had a licence to blink. 
Indeed, you will notice that if you change the linearity setting down in SVDecon, the ringing becomes less.
for helping to crack the case);