Originally posted by UserAccessDenied
EDIT:
I just realized my Fiancee has artrage on her laptop. It's not the greatest for photo editing, but I can use a selective tool and overlap the PS/non-PS image like you mention and erase the artifacts to reveal the single non-PS image underneath.
Still not an ideal workflow, but it's the object is moving like a long-exposure on a highway or flowing water, what do you expect PS to do with it anyhow?
It's not like we want a sharper blur? or do we?
There's desirable blur (think bokeh, depth of field, motion blur, star trails, etc) and there are crappy artifacts (think dust motes, scratches on film, red eye, and yes, PS artifacts).
You can still get blur in PS images (it's like a bracket of four shots slightly moved). So each of the four shots has blur, which can be not so bad. But on the edges you can get these lattices ugly colored highly artificial blotches. Over in other threads you can see them.
And unlike other blur, it's hard to fix in the field as well. I was taking macro shots of flowers the other day, and even the tiniest air movement on little filaments and pollen and petals caused artifacts, and since that's the main subject, PS was worthless. And this is in bright light; no amount of shutter speed can fix it, unlike regular motion blur. The very marginal increase in sharpness just isn't worth it in many cases in the real world. In studio work or maybe very still landscapes you can make it work, but even then it would be with an eye toward large prints and other output that could make use of it, not posting online (which is why you don't see that many example posted online).
PS reminds me of focus stacking and similar techniques. Handy to have, but a bit of work, and something you usually don't resort to if you can get what you want in other ways, like just stopping down and adding light in the case of focus stacking. Even if there were easy fixes in post processing that didn't mess up parts of the photo it still isn't something that's a substitute for perhaps a better lens choice or other techniques that are more flexible. But hey, panorama processing used to be really iffy and now it works amazingly well, so perhaps in time it will improve. I applaud having options like it. There might even be things to do with the unique aspects of sensor movement, as Astrotracer does, like a sensor motion pan, or even more frames over a wider area. Who knows.